Nurk E, Refsum H, Drevon CA, et al. Intake of flavonoid-rich wine, tea, and chocolate by elderly men and women is associated with better cognitive test performance. J Nutr. 2009;139: 1-8.

Neurodegenerative disorders, particularly Alzheimer's disease, progressively lead to cognitive impairment and dementia. Environmental and genetic factors appear to contribute equally to the pathogenesis of neurodegenerative disorders, and the role of micronutrients in cognitive ability has become a topic of increasing interest. Polyphenols, which are powerful antioxidants, are abundant in plant-derived foods. Fruits, vegetables, chocolate, red wine, and tea are rich sources of flavonoids—the largest subclass of polyphenols. Studies have shown an inverse relation between cognitive performance and the intake of flavonoids. Short-term and long-term consumption of cocoa has been shown to be associated with an increase in blood flow to the cerebral gray matter, which is beneficial in patients with conditions involving a reduction in cerebral blood flow. The objective of the present study was to assess the relation between habitual intake of flavonoid-rich foods and drinks (i.e., chocolate [Theobroma cacao], wine [Vitis vinifera], and tea [Camellia sinensis]) and cognitive performance in elderly men and women.

A subset of elderly men and women from the Hordaland Health Study (conducted in Oslo, Norway) were recruited to undergo a series of cognitive tests. The cross-sectional analysis was limited to 2031 adults (55% women) aged 70-74 years. The subjects completed a food-frequency questionnaire designed to elicit information on the intake of 169 foods over the past year. The intakes of chocolate, wine, and tea were calculated for each subject, and the subjects were then classified as consumers or nonconsumers of these items. After undergoing routine cardiovascular examination, the subjects completed 6 cognitive tests: the Kendrick Object Learning test (KOLT); the Trail Making test, part A (TMT-A); a modified version of the Digital Symbol test (m-DST); the Block design, short form (m-BD); a modified version of the Mini-Mental State Examination (m-MMSE); and an abridged version of the Controlled Oral Word Association test (S-task). These tests measure episodic memory, cognitive level and memory, perceptual speed, visuospatial skills, global cognition, and semantic memory, respectively. Poor cognitive performance was defined as a score in the lowest decile for all of the tests except the TMT-A, for which a score in the highest decile indicated poor cognitive performance.

The mean intakes of chocolate, wine, and tea were 3.8 g/d, 22 mL/d, and 222 mL/d, respectively, for the total group of 2031 participants. The respective mean intakes doubled when only the "consumers" were considered: 7.5 g/d (n = 1036), 51 mL/d (n = 882), and 417 mL/d (n = 827) [Editor's Note: This number (827) is given in the first paragraph under Results; however, Table 1 indicates that the number is 1083 which is more likely the accurate number (combined with 948 nonconsumers to total 2031).]. More men than women drank wine, and more women than men drank tea. Chocolate consumers had a lower prevalence of cardiovascular disease history and diabetes than did nonconsumers of chocolate, while diabetes was lower among wine drinkers than nondrinkers. Consumers of chocolate, wine, or tea scored better on all 6 cognitive performance tests than did nonconsumers of these foods (P<0.001, except P=0.015 on KOLT score for tea). Furthermore, the performance on all 6 tests improved with increasing wine intakes up to 75–100 mL/d and with increasing intakes of chocolate up to about 10 g/d for all tests except KOLT. The greatest dose-response effect observed for tea was at an intake of about 200 mL/d. Consumers of all 3 test food substances had the best cognitive performance scores.

The habitual intake of the flavonoid-rich foods chocolate, wine, and tea were associated with better cognitive performance, and the associations for each of these foods were dose-dependent for all of the tests with the exception of the association between chocolate intake and performance on the KOLT. The associations were also additive, i.e., the risk of poor cognitive performance decreased as the intake of the 3 flavonoid-rich foods increased. The major strength of this study was that it included a large population-based sample and that 6 different tests were used to assess cognitive performance. One limitation of the study was its use of a food-frequency questionnaire, which often over- or underestimate actual intakes, since it did not distinguish between dark and milk chocolate, red and white wine, or green and black tea; however, the food-frequency questionnaire used in the present study was validated in several other studies. The authors conclude that the "intake of flavonoid-rich food, including chocolate, wine, and tea, is associated with better performance across several cognitive abilities and that the associations are dose dependent." However, they recommend that the results be considered "with caution" because they are based on food-based analyses of a population.