Farzana Jahan Phd, FHEA

Background
Alzheimer's disease (AD), the most common form of dementia, is marked by significant reductions in glucose metabolism. Such hypometabolism reflects underlying synaptic dysfunction, correlating with cognitive decline. Our study aimed to explore the impact of dietary patterns—specifically, the Western Diet and Prudent Diet—on change in glucose metabolism in brain regions associated with AD risk, [18F]Fluorodeoxyglucose positron emission tomography (FDG‐PET) imaging as a biomarker.

Method
Longitudinal data from 133 cognitively unimpaired older adults were analysed from the Western Australian Memory Study. Participants underwent dietary assessment using the Cancer Council of Victoria Food Frequency Questionnaire and completed FDG‐PET imaging up to three times over 43 months. Dietary patterns were identified through principal component analysis, yielding two patterns—named Western Diet and Prudent Diet. Pattern scores were computed by summing food group intakes weighted by their respective factor loadings. Linear mixed‐effect models evaluated the association between dietary adherence and brain glucose metabolism, including potential confounders. The cohort was stratified by apolipoprotein E (APOE) ε4 carrier status, a genetic risk factor for AD, to investigate potential differing effects.

Result
Adherence to a Western Diet, characterised by high sugars and saturated fats, was associated with a faster decline in glucose metabolism in the right fusiform gyrus among APOE ε4 carriers (β = ‐0.00012; SE = 0.00004; false discover rate adjusted p = .032), with no significant associations in APOE ε4 non‐carriers. Similarly, no significant associations were observed between the Prudent Diet, characterised by high intake of fruits, vegetables, and whole grains, and glucose metabolism, in both APOE ε4 carriers and non‐carriers.

Conclusion
Our study highlights the potential detrimental impact of a Western Diet on brain glucose metabolism, particularly for individuals at genetic risk for AD. The decline in glucose metabolism in the fusiform gyrus, a region essential for cognitive functions like facial recognition, emphasises the role of diet in brain health. Future research should examine the mechanisms linking diet to neurodegeneration and explore dietary interventions as preventive strategies against cognitive decline and dementia.

To find the best model for different link functions in Generalized Linear Models has been a challenge to researchers and users. Lindsey and Jones (1988) provided some procedures to suggest the techniques for choosing among the generalized linear models applied to medical data. They employed Akaike's Information Criteria (AIC). In the present study, an attempt is made to compare the performance in choosing the best GLM using the alternative procedures such as: deviance. Simulation from different distributional assumptions are performed to demonstrate the comparison mong the alternative approaches for various link functions. In addition to existing methods based on distributional assumptions, an alternative approach is suggested based on the quasi likelihood approach where no distributional assumptions is needed. The suggested methods for comparison of models are also applied to the same data set which Lindsey and Jones (1998) used in their study; data on T4 cells/mm3 in blood samples from 20 patients in remission from Hodgkins disease and 20 other patients in remission from disseminated malignancies. The performance of deviance appears to be satisfactory for this data.