Abstract: | Objectives: To find dietary patterns and foods associated with cognitive function. Design: A cross-sectional study of short-term effects and a prospective study for long-term effects. Setting: Nutrition and Health Survey in Taiwan (NAHSIT) 2014-2016 and NAHSIT 1999-2000. Participants: A total of 1245 older patients enrolled in the NAHSIT 2014-2016 and 1436 in the NAHSIT 1999-2000. Measurements: Dietary intake was appraised with a food-frequency questionnaire. Cognitive function was assessed by the Mini-Mental State Examination score (MMSE). Results: Using reduced rank regression to data-mine NAHSIT 2014-2016 cross-sectional data, we found in both genders a dietary pattern associated with high MMSE score, which was characterized by higher intakes of fresh fruits, nuts and seeds, whole grains, breakfast cereals, coffee, dairy products, seafood products, and fish. Moreover, in women, the pattern included a few additional items: tea, eggs, soybean products, and vegetables. Presence of mild cognitive impairment was inversely associated with the dietary pattern score, with declined adjusted odds ratio (95% confidence interval) from tertile 1 (as reference), tertile 2, to tertile 3 in both men [1 → 0.85(0.45-1.61) → 0.32 (0.14-0.78)] and women [1→0.44 (0.25-0.76) → 0.39 (0.20-0.75)]. Using the NAHSIT 1999-2000 as a baseline, along with 11 years of follow-up, we found with the Cox proportional hazards model that higher intake (≥4 vs <1 time/wk) of either tea or fish, but not other foods, was associated with a lower risk of developing dementia. Higher intakes of both tea and fish were associated with an even lower risk. Conclusions/Implications: A dietary pattern characterized by high intakes of phytonutrient-rich plant foods (fruits, whole grains, nuts/seeds, and vegetables), tea and coffee, and protein-rich foods such as eggs, dairy products, and fish, was associated with the presence of better cognitive function in older adult. Higher intakes of fish and tea combined showed a long-term protective effect. Further research is warranted to understand the long- and short-term effects of diet. |