Alzheimer's disease and diet: a systematic review

Abstract

Purpose/Aim: Approximately 44 million people worldwide have Alzheimer's disease (AD). Numerous claims have been made regarding the influence of diet on AD development. The aims of this systematic review were to summarize the evidence considering diet as a protective or risk factor for AD, identify methodological challenges and limitations, and provide future research directions. Methods: Medline, PsycINFO and PsycARTICLES were searched for articles that examined the relationship between diet and AD. Results: On the basis of the inclusion and exclusion criteria, 64 studies were included, generating a total of 141 dietary patterns or “models”. All studies were published between 1997 and 2015, with a total of 132 491 participants. Twelve studies examined the relationship between a Mediterranean (MeDi) diet and AD development, 10 of which revealed a significant association. Findings were inconsistent with respect to sample size, AD diagnosis and food measures. Further, the majority of studies (81.3%) included samples with mean baseline ages that were at risk for AD based on age (>65 years), ranging from 52.0 to 85.4 years. The range of follow-up periods was 1.5–32.0 years. Conclusions: The mean age of the samples poses a limitation in determining the influence of diet on AD; given that AD has a long prodromal phase prior to the manifestation of symptoms and decline. Further studies are necessary to determine whether diet is a risk or protective factor for AD, foster translation of research into clinical practice and elucidate dietary recommendations. Despite the methodological limitations, the finding that 50 of the 64 reviewed studies revealed an association between diet and AD incidence offers promising implications for diet as a modifiable risk factor for AD.

KEYWORDS:

• Alzheimer's
• diet
• protective
• risk
• food

Acknowledgements

We thank Morris Byk for his comments on an earlier version of this paper.

Declaration of Interests

Authors M. Yusufov, L. L. Weyandt and I. Piryatinsky report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.