PROJECT SUMMARY / ABSTRACT Diet is a promising modifiable factor for the prevention of late-onset Alzheimer's disease and related dementias (LOADRD). Healthy diet may directly reduce oxidative stress and inflammation in the brain and moderate metabolic risk factors of LOADRD, such as obesity, diabetes, and heart disease. Specific foods and nutrients (e.g., vegetables, fat-soluble vitamins, omega-3 fatty acids, polyphenols) have been suggested as protective. However, considering the complexity of diet and the synergistic effects of constituents, increasingly the research focus is on overall dietary patterns: especially for LOADRD research, the Mediterranean diet (MED), the Dietary Approaches to Stop Hypertension (DASH) and the Mediterranean-DASH intervention for Neurodegenerative Delay (MIND) diet. While some prospective observational studies provide strong support for the protective effect of these dietary patterns, mostly involving White populations and some African Americans, data is lacking for Latino American, Asian American, and Native populations, which limits the understanding of dietary contributions to ADRD disparities. Also, longitudinal studies of dietary change or maintenance are needed to provide evidence for dietary interventions to reduce ADRD risks. To address these gaps, we propose a secondary data analysis in the Multiethnic Cohort (MEC) Study on African American, Japanese American, Latino, Native Hawaiian, and White adults with high-quality dietary data and Medicare linkage-based ADRD outcomes (1999-2019). Based on a validated food frequency questionnaire administered at cohort entry and 10-year follow-up, we have calculated several dietary pattern scores (the alternate MED (aMED), DASH, the Healthy Eating Index 2015 (HEI-2015), and the Dietary Inflammatory Index (DII)) and will compute the MIND score. Our analysis will include ~102,000 participants and ~19,000 ADRD cases. Specifically, we will evaluate the LOADRD risk associated with overall dietary patterns (MIND, aMED, DASH, HEI-2015 and DII scores) (Aim 1) and the individual components of the patterns and two specific nutrients (fat-soluble vitamins and polyphenols) (Aim 2) overall and by race and ethnicity. We will assess these associations (a) based on the baseline intake, (b) for the change over 10 years, and (c) by ADRD subtypes and by genetic risk (APOE e4 status or polygenic risk score). We hypothesize that healthful dietary patterns will be protective against LOADRD across racial and ethnic groups, that improved dietary quality will be associated with a reduced risk, and that the specific bioactive components, the consumption of which vary by race and ethnicity, may have independent protective effects. Lastly, we will develop a novel dietary pattern to capture the unique and diverse dietary features of the multiethnic population, and validate it by examining its association with LOADRD (Aim 3). This project based in a large, long-term, multiethnic cohort, equipped wi...