One cognitive ability stood out in a study of Western diets and Alzheimer’s risk genes

By Vijay Kumar MalesuReviewed by Susha Cheriyedath, M.Sc.Aug 11 2026

Long-term Australian data suggest Western eating patterns may be linked to faster attention decline in genetically susceptible older adults, but sensitivity analyses show the association remains uncertain.

Study: Dietary Patterns and Cognitive Decline in Older Adults: Findings from the Western Australia Memory Study. Image Credit: Yuriy Golub / Shutterstock

In a recent study published in the journal Nutrients, a group of researchers in Australia examined whether Western and Prudent dietary patterns are associated with longitudinal cognitive change according to apolipoprotein E (APOE) ε4 carrier status in older adults.

Background

As people live longer, how can everyday food choices support cognitive health? Global life expectancy increased by more than six years from 2000 to 2019, while Alzheimer’s disease (AD) has become an important age-related condition.

Cognitive decline can affect memory, language, attention, and executive function and may progress to dementia. Dietary patterns have attracted interest because they represent combinations of foods and nutrients consumed routinely.

Previous research has produced inconsistent findings, likely due to differences in dietary assessment, cognitive outcomes, follow-up duration, and genetic susceptibility. Further research is needed using long-term, domain-specific cognitive assessments and analyses that consider APOE genotype.

About the study

Longitudinal data from 185 adults in the Western Australia Memory Study (WAMS) were analyzed. Participants were recruited in Perth from 2006 to 2009. They provided informed consent, completed a neuropsychological battery and the Cancer Council of Victoria Food Frequency Questionnaire (CCVFFQ), and donated blood. Baseline exclusions included uncontrolled psychiatric disorders, psychotic disorders, stroke, and neurological diseases affecting cognition.

Cognitive assessments were conducted at baseline and at up to 5 visits at 18-month intervals. Six cognitive domains were assessed, with scores combined into composites and a Preclinical Alzheimer’s Cognitive Composite (PACC).

Dietary intake was assessed using the CCVFFQ. Principal component analysis (PCA) of 33 food groups identified Western and Prudent dietary patterns. APOE genotyping was performed from blood-derived deoxyribonucleic acid (DNA).

Analyses used R version 4.4.1 and Statistical Package for the Social Sciences (SPSS) version 30. Linear regression assessed associations, while linear mixed-effects models assessed cognitive change over up to 90 months.

Overall, cross-sectional and longitudinal models were adjusted for age, sex, education, body mass index (BMI), energy intake, and APOE genotype. APOE ε4-stratified cross-sectional models adjusted for age, sex, and education, while stratified longitudinal models adjusted for age, sex, education, BMI, and energy intake. The analysis was corrected for false discovery rate (FDR), with p-value < 0.05 considered significant.

Furthermore, a sensitivity analysis was conducted by excluding those participants who scored below 23 points on the Montreal Cognitive Assessment (MoCA) test to identify possible baseline cognitive impairment; this cutoff was not intended as a diagnostic threshold.

Study results

The cohort included 185 older adults with a mean age of 70.6 ± 6.0 years; 31.9% were men, the mean BMI was 26.6 ± 4.2 kg/m², and 30.8% carried an APOE ε4 allele. More than one-quarter had 12 or fewer years of education. APOE ε4 carriers had a higher proportion of men and a lower BMI than non-carriers, but dietary pattern scores did not differ significantly between groups.

A total of 36 participants completed the sixth cognitive assessment, and 19 had baseline MoCA scores below 23. This group did not differ significantly in age, BMI, energy intake, Western diet score, Prudent diet score, or APOE ε4 carriage. However, the group with scores below 23 had a higher proportion of men and fewer years of education.

During follow-up, 22 participants developed one or more cardiometabolic conditions. Those participants had higher Western diet scores and fewer years of education, while other baseline characteristics did not differ significantly. Additional analyses incorporating cardiometabolic conditions as a potential confounder did not alter the primary findings.

Cross-sectional linear regression analyses found no significant associations between either dietary pattern and cognitive performance after FDR adjustment. No significant associations were identified among APOE ε4 carriers or non-carriers after correction for multiple comparisons. Thus, baseline dietary pattern scores were not significantly related to cognitive performance in these analyses.

Longitudinal linear mixed-effects analyses also found no significant associations between dietary patterns and cognitive decline in the overall cohort. After FDR adjustment, higher Western diet adherence was significantly associated with faster decline in the attention composite among APOE ε4 carriers (β = −0.0028; standard error [SE] = 0.00095; FDR-adjusted p = 0.016).

The partial eta squared for this association was 0.18. However, the authors cautioned that the magnitude of the effect was imprecisely estimated and that its clinical significance remains uncertain. 

No significant associations were observed between Western diet adherence and other cognitive domains in APOE ε4 carriers, and no significant associations were observed among APOE ε4 non-carriers. The Prudent dietary pattern was not significantly associated with cognitive decline.

The significant association between the Western diet and attention decline in APOE ε4 carriers was attenuated after excluding participants with possible baseline cognitive impairment (MoCA score < 23). In that sensitivity analysis, the association was no longer statistically significant (β = −0.0017; SE = 0.00092; FDR-adjusted p = 0.233).

The researchers, therefore, characterized the significant longitudinal finding as exploratory. They also reported that genotype-stratified analyses were underpowered, particularly at the final assessment because of attrition, and that the null finding for the Prudent diet was inconclusive rather than definitive.

Conclusions

Among older adults in the WAMS, adherence to a Western dietary pattern was associated with faster decline in attention among APOE ε4 allele carriers, but this association weakened and lost statistical significance after participants with possible baseline cognitive impairment (MoCA score < 23) were excluded. No statistically significant cognitive associations were observed for the Prudent dietary pattern or among APOE ε4 non-carriers.

The findings suggest a possible relationship between dietary patterns and genetic susceptibility in aging, but the association was exploratory, with subgroup sizes and attenuation in sensitivity analysis. Diet was assessed only at baseline using a self-reported food frequency questionnaire, and the authors noted that reverse causality could not be excluded. Longitudinal cohorts and intervention studies are needed before recommendations can be made.