From the Institute for Risk Assessment Sciences (Y.Z., H.K., S. Peters, R.V.), Utrecht University, the Netherlands; Department of Environmental Health Sciences (Y.L., G.W.M.), Mailman School of Public Health, Columbia University, New York, NY; Department of Epidemiology (J.M.H.), Murcia Regional Health Council-IMIB, Murcia; CIBER Epidemiología y Salud Pública (CIBERESP) (J.M.H., M.G.), Madrid; Movement Disorders Unit (A.V.-A.), Department of Neurology, University Hospital Donostia; BioDonostia Health Research Institute (A.V.-A.), Neurodegenerative Diseases Area, San Sebastián, Spain; Division of Cancer Epidemiology (J.A.S.), German Cancer Research Center (DKFZ), Heidelberg, Germany; Danish Cancer Institute (J.H.), Danish Cancer Society, Copenhagen, Denmark; Escuela Andaluza de Salud Pública (EASP) (D.P.); Instituto de Investigación Biosanitaria-ibs.GRANADA (D.P.), Granada; Centro de Investigación Biomédica en Red de Epidemiología y Salud Pública (CIBERESP) (D.P.), Madrid, Spain; Unit of Cancer Epidemiology (C.S.), Città della Salute e della Scienza University-Hospital, Turin, Italy; Unit of Nutrition and Cancer (R.Z.-R.), Cancer Epidemiology Research Programme, Catalan Institute of Oncology (ICO), Bellvitge Biomedical Research Institute (IDIBELL), Barcelona, Spain; Epidemiology and Prevention Unit (V.P.), Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Italy; Department of Epidemiology and Biostatistics (A.K.H., M.G.), School of Public Health, Imperial College London, United Kingdom; School of Medicine (S. Panico), Federico II University, Naples, Italy; de Salud Pública y Laboral de Navarra (M.G.), Pamplona; Navarra Institute for Health Research (IdiSNA) (M.G.), Pamplona, Spain; Institute for Cancer Research (G.M.), Prevention and Clinical Network (ISPRO), Florence, Italy; Institute of Epidemiology and Social Medicine (C.M.L.), University of Münster, Germany; Ageing Epidemiology Research Unit (AGE) (C.M.L.), School of Public Health, Imperial College London, United Kingdom; and University Medical Centre Utrecht (R.V.), the Netherlands.
Neurology. 2024 Apr 23;102(8):e209201. doi: 10.1212/WNL.0000000000209201. Epub 2024 Mar 21.
Inverse associations between caffeine intake and Parkinson disease (PD) have been frequently implicated in human studies. However, no studies have quantified biomarkers of caffeine intake years before PD onset and investigated whether and which caffeine metabolites are related to PD.
Associations between self-reported total coffee consumption and future PD risk were examined in the EPIC4PD study, a prospective population-based cohort including 6 European countries. Cases with PD were identified through medical records and reviewed by expert neurologists. Hazard ratios (HRs) and 95% CIs for coffee consumption and PD incidence were estimated using Cox proportional hazards models. A case-control study nested within the EPIC4PD was conducted, recruiting cases with incident PD and matching each case with a control by age, sex, study center, and fasting status at blood collection. Caffeine metabolites were quantified by high-resolution mass spectrometry in baseline collected plasma samples. Using conditional logistic regression models, odds ratios (ORs) and 95% CIs were estimated for caffeine metabolites and PD risk.
In the EPIC4PD cohort (comprising 184,024 individuals), the multivariable-adjusted HR comparing the highest coffee intake with nonconsumers was 0.63 (95% CI 0.46-0.88, = 0.006). In the nested case-control study, which included 351 cases with incident PD and 351 matched controls, prediagnostic caffeine and its primary metabolites, paraxanthine and theophylline, were inversely associated with PD risk. The ORs were 0.80 (95% CI 0.67-0.95, = 0.009), 0.82 (95% CI 0.69-0.96, = 0.015), and 0.78 (95% CI 0.65-0.93, = 0.005), respectively. Adjusting for smoking and alcohol consumption did not substantially change these results.
This study demonstrates that the neuroprotection of coffee on PD is attributed to caffeine and its metabolites by detailed quantification of plasma caffeine and its metabolites years before diagnosis.
在人类研究中，经常提到咖啡因摄入与帕金森病（PD）之间呈负相关。然而，尚无研究对 PD 发病前数年的咖啡因摄入量进行生物标志物定量分析，并探讨哪些以及何种咖啡因代谢物与 PD 相关。
在 EPIC4PD 研究中，对来自 6 个欧洲国家的前瞻性人群队列进行了自我报告的总咖啡摄入量与未来 PD 风险的相关性研究。通过病历和专家神经科医生的审查来确定 PD 病例。使用 Cox 比例风险模型估计咖啡摄入量与 PD 发生率的风险比（HR）和 95%置信区间（CI）。在 EPIC4PD 中嵌套进行了病例对照研究，招募了新发 PD 病例，并通过年龄、性别、研究中心和采血时的禁食状态与每个病例匹配对照。使用高分辨率质谱法在基线采集的血浆样本中定量咖啡因代谢物。使用条件逻辑回归模型估计了咖啡因代谢物与 PD 风险的比值比（OR）和 95%CI。
在 EPIC4PD 队列（包含 184024 人）中，最高咖啡摄入量与非消费者相比的多变量调整 HR 为 0.63（95%CI 0.46-0.88， = 0.006）。在包括 351 例新发 PD 病例和 351 例匹配对照的嵌套病例对照研究中，诊断前的咖啡因及其主要代谢物，即黄嘌呤和茶碱，与 PD 风险呈负相关。OR 分别为 0.80（95%CI 0.67-0.95， = 0.009）、0.82（95%CI 0.69-0.96， = 0.015）和 0.78（95%CI 0.65-0.93， = 0.005）。调整吸烟和饮酒因素后，这些结果并未发生实质性改变。
本研究通过对诊断前数年的血浆咖啡因及其代谢物进行详细定量分析，证明了咖啡对 PD 的神经保护作用归因于咖啡因及其代谢物。
