INTRODUCTION Long-term use of proton pump inhibitors (PPIs) has been associated with a wide variety of potentially serious adverse effects including a possible increased risk of dementia. Studies evaluating this association have reached divergent conclusions. We aimed to evaluate this proposed association further and to assess the quality of the evidence in its support. METHODS We searched MEDLINE, EMBASE, ISI Web of Science, and Cochrane databases for studies examining a link between PPI use and dementia, up to February 2019. Studies reporting summary results as hazard ratio (HR) or odds ratio (OR) were pooled using the DerSimonian and Laird random-effects model for meta-analyses. Methodological quality of individual observational studies was assessed using the Newcastle-Ottawa scale and the overall quality of evidence rated as per the GRADE approach. RESULTS We identified and included 11 observational studies comprising 642,949 subjects; 64% were women. Most studies were short-term ranging from 5 to 10 years. There were 158,954 PPI users and 483,995 nonusers. For studies summarizing data as adjusted HR, pooled HR for all causes of dementia was 1.10 (0.88-1.37); for Alzheimer dementia only, it was 1.06 (0.72-1.55). For studies summarizing data as adjusted OR, pooled OR for all causes of dementia was 1.03 (0.84-1.25) and for Alzheimer dementia only 0.96 (0.82-1.11). Per Newcastle-Ottawa scale assessment, 10 studies were of high quality and 1 was of moderate quality. By applying GRADE methodology, quality of evidence for both outcomes was very low. DISCUSSION We found no evidence to support the proposed association between PPI use and an increased risk of dementia. PPI use among patients who have a valid indication for it, should not be curtailed because of concerns about dementia risk.

中文翻译：

---

短期使用质子泵抑制剂与痴呆症没有关联

引言 长期使用质子泵抑制剂 (PPI) 会导致多种潜在的严重不良反应，包括可能增加患痴呆症的风险。评估这种关联的研究得出了不同的结论。我们旨在进一步评估这一提议的关联，并评估支持其的证据质量。方法 我们在 MEDLINE、EMBASE、ISI Web of Science 和 Cochrane 数据库中搜索了研究 PPI 使用与痴呆之间联系的研究，截至 2019 年 2 月。将汇总结果报告为风险比 (HR) 或优势比 (OR) 的研究使用汇总用于元分析的 DerSimonian 和 Laird 随机效应模型。使用纽卡斯尔-渥太华量表评估个别观察性研究的方法学质量，并根据 GRADE 方法对证据的整体质量进行评级。结果 我们确定并纳入了 11 项观察性研究，包括 642,949 名受试者；64% 是女性。大多数研究是短期的，从 5 年到 10 年不等。有 158,954 名 PPI 用户和 483,995 名非用户。对于将数据汇总为调整后的 HR 的研究，所有痴呆原因的汇总 HR 为 1.10 (0.88-1.37)；仅对于阿尔茨海默痴呆症，它是 1.06 (0.72-1.55)。对于将数据汇总为调整后的 OR 的研究，所有痴呆原因的汇总 OR 为 1.03 (0.84-1.25)，而阿尔茨海默痴呆仅为 0.96 (0.82-1.11)。根据纽卡斯尔-渥太华量表评估，10 项研究为高质量，1 项为中等质量。通过应用 GRADE 方法，这两种结果的证据质量都非常低。讨论 我们发现没有证据支持 PPI 使用与痴呆风险增加之间的拟议关联。不应因为担心痴呆风险而在有有效适应症的患者中使用 PPI。