Date of Award


Degree Type


Degree Name

Doctor of Philosophy


School of Medical Sciences


Health, Engineering and Science

First Advisor

Professor Ralph Martins

Second Advisor

Doctor Stephanie Rainey-Smith

Third Advisor

Associate Professor Jennifer Keogh

Fourth Advisor

Mr Kevin Taddei


Alzheimer’s disease (AD), the most common form of senile dementia, currently affects over 35 million people worldwide. While there is no cure or effective treatment, early intervention programs hold considerable promise. Following particular dietary patterns represents one potential intervention strategy accessible to all.

Results from previous studies investigating the association of diet, cognition and biomarkers of AD are inconsistent: Positive results have been reported (1-7), whilst others have shown no associations. Prior to this thesis, no study has assessed the relationship of four dietary patterns to cognition, blood-based and neuroimaging biomarkers of AD in a large highly-characterised ageing cohort. Participants drawn from the Australian Imaging, Biomarkers, and Lifestyle study of ageing, provided a fasting blood sample, underwent comprehensive neuropsychological assessment and neuroimaging at baseline, 18 and 36 month follow-up assessments, and completed a Cancer Council of Victoria food frequency questionnaire (used to construct dietary patterns) at baseline.

Chapter 3 explored the relationship between dietary pattern adherence and cognition. AD participants demonstrated reduced adherence to the ‘healthy’ Mediterranean (MeDi) and prudent diets, and higher adherence to the ‘unhealthy’ western diet and the inflammatory dietary index compared to cognitively healthy controls (HC). Longitudinal analysis conducted on individuals classified as HC at baseline proposes the importance of adhering to a ‘healthy’ dietary pattern such as the MeDi, with respect to reducing risk for cognitive decline: Executive function and visuospatial functioning appeared most susceptible to the influence of diet.

Chapter 4 investigated the potential mechanisms underlying the observed effects of dietary patterns on cognition. A lack of significant associations between the MeDi and western diet patterns and biomarker indexes of metabolic syndrome and cardiovascular disease risk, suggests that modulation of these factors may not underlie the effects of diet on cognition reported in Chapter 3.

Consistent with published literature, we found our western dietary pattern to be positively associated with levels of blood-based biomarkers of inflammation and the reverse to be true of our MeDi and prudent diet patterns. Our inflammatory dietary index was also strongly positively correlated with levels of numerous inflammatory biomarkers. The strong associations observed suggest that interplay between diet and elevated chronic inflammation may contribute to the effects of diet on cognition described in this thesis.

Chapter 5 assessed the ‘reliability’ (similarity of 12 month dietary intake recalled on different occasions) and ‘validity’ (intake agreement between FFQ and a four-day weighed food record) of the online CSIROFFQ following addition of questions regarding foods of interest in AD research. Our results suggest that the modified CSIROFFQ is ‘reliable’ and a ‘relatively valid’ tool which provides acceptable assessment of long-term dietary intake in Australian older adults, particularly in the context of AD research.

To our knowledge, this is the first study extensively comparing MeDi, inflammatory dietary index, western and prudent diet patterns to cognition and biomarkers of AD in an elderly, well-characterised cohort. Our results combined with published data, suggest diet has a role to play in AD prevention; however, it is clear that the complex link requires further characterisation.