Associations between diet, cognitive function and dementia risk in UK adults

Cognitive decline and dementia are of increasing concern in aging societies worldwide. Diet, as a modifiable lifestyle factor, represents a target for prevention or limiting progression. However, evidence on associations of cognitive function and dementia with diet remains limited and inconsistent, especially on meat consumption summarized in the systematic review of this project. Cross-sectional associations of dietary factors with one cognitive test (reaction time) and dementia (ascertained via death registers) were conducted in UK Women’s Cohort Study (UKWCS). The results showed that consumption of specific food groups, energy-adjusted nutrient intakes, and adherence to dietary patterns were not statistically associated with reaction time and dementia in the UKWCS. Cross-sectional and longitudinal associations of food consumption, especially meat intakes, with five cognitive tests (visual memory, numeric memory, prospective memory, fluid intelligence, and reaction time) and dementia (ascertained via self-report and linkages to hospital admission data and death registers) were conducted in UK Biobank (UKB). Incident dementia cases occurring within 1-year or 3-year follow-up were excluded due to potential reverse causation, and similar results were observed between the two types of exclusion. The results showed that high consumption of processed meat was associated with increased risks of prevalent and incident dementia; with a non-linear pattern of this association indicated in the UKB. Associations between consumption of other meat types and cognitive performance and dementia risk were not consistent in the UKB. A diet-gene interaction of APOE ε4 allele on dementia risk was explored, and all P values for interaction were not significant. In addition, high consumption of vegetables, fruits, and fish were observed to be associated with poor cognitive performance and increased risk of incident dementia in the UKB although effect sizes were small. This project highlights potentially non-linear associations between meat consumption and dementia risk, which may be independent of APOE ε4 allele carriage. Findings on consumption of vegetables, fruits, and fish were not consistent with the hypotheses proposed of a protective effect in this thesis. However, the effect sizes were relatively small and therefore need to be interpreted with caution and to be confirmed in other studies.