Teuku Fais Duta, Derren D. C. H. Rampengan, Nuril Farid Abshori, Bryan Gervais de Liyis, Fahrul Nurkolis, Muhammad Iqhrammullah
Background: Observational links between olive oil and lower dementia risk may reflect confounding or survival bias. We tested whether olive oil consumption causally influences Alzheimer’s disease (AD) risk using Mendelian randomization (MR). Methods: We performed two-sample MR using IEU OpenGWAS summary statistics. Genetic instruments for olive oil consumption were derived from a UK Biobank cooking-fat question (P < 5×10-6, LD-independent), explaining around 0.07% of exposure variance with mean F of around 24. The AD outcome was a large Europeanancestry GWAS including clinically diagnosed and proxy cases. The primary estimator was inverse-variance weighted (IVW), with MR-Egger, Cochran’s Q and leave-one-out analyses for sensitivity. Two-step MR tested mediation via lipid traits (LDL-C, HDL-C, triglycerides, apolipoprotein B, apolipoprotein A1), blood pressure and inflammatory markers. Multivariable MR (MVMR) adjusted for apolipoprotein B, adiposity, blood pressure, and systemic inflammation. Results: There was no evidence that higher genetically proxied olive oil intake reduces AD risk. The IVW point estimate was effectively null, with confidence intervals excluding even modest benefits (for example, > 1% relative risk reduction per SD increase). Findings were robust across MR-Egger, heterogeneity tests and leaveone-out analyses, with no indication of horizontal pleiotropy. Mediation analyses showed no indirect effects through lipid profiles, vascular injury or inflammation. In MVMR, the olive oil coefficient remained close to null, and the adjusted effects of other traits were non-significant. Conclusion: This comprehensive MR analysis does not support a causal neuroprotective effect of olive oil on AD. Any true effect, if present, is likely minimal. Observational associations may reflect confounding or healthy-user bias. While cardiovascular benefits justify olive oil as part of a heart-healthy diet, it should not be promoted as a stand-alone AD prevention strategy. Larger, ancestry-diverse datasets and gene–diet interaction analyses are warranted. © 2025, Ant Publishing Corporation. All rights reserved.
Medical Research Unit, Faculty of Medicine, Universitas Syiah Kuala, Banda Aceh, Indonesia; Faculty of Medicine, Universitas Sam Ratulangi, Manado, Indonesia; Faculty of Medicine and Health Sciences, Maulana Malik Ibrahim Islamic State University Malang, Malang, Indonesia; Department of Neurosurgery, National Brain Center Hospital Prof. Dr. dr. Mahar Mardjono, Jakarta, Indonesia; Master of Basic Medical Science, Faculty of Medicine, Universitas Airlangga, Surabaya, Indonesia; Medical Research Center of Indonesia, Surabaya, Indonesia; State Islamic University of Sunan Kalijaga (UIN Sunan Kalijaga), Yogyakarta, Indonesia; Postgraduate Program of Public Health, Universitas Muhammadiyah Aceh, Banda Aceh, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock