The role of dietary choices in cognitive behaviour: a prospective population-based study

Abstract

BackgroundCognitive decline is the first outward sign of dementia and has a major public health effect. Evidence suggests that dietary patterns are associated with cardiovascular and cerebrovascular risk factors. A high quality diet or Mediterranean diet might protect against Alzheimer's disease, but its association with age-associated cognitive decline has not been fully explored. We aimed to examine the association between two ways of quantifying dietary patterns (a healthy dietary scale and a Mediterranean diet) and 20 years of cognitive decline from age 43 years to 60–64 (>60) years in a prospective longitudinal population-based study. MethodsWe included 1018 participants from the MRC National Survey of Health and Development (the British 1946 birth cohort) regularly followed up over 23 times across their entire life since birth in 1946. Cognitive functioning was measured at age 43 years and older than 60 years by tests of verbal memory (maximum words recalled 45) and timed visual search (maximum letter searched 600). The two dietary patterns were constructed with information from a 5 day estimated diet diary at both 36 and 43 years. The healthy dietary scale represents an overall score of healthy food choices: frequency of breakfast; type of milk; type of bread; number of daily portions of fruit and vegetables; and average percentage of energy from fat. The total score was dichotomised by median split at each age to represent lower or higher quality diet. An adherence scale to the Mediterranean diet was also derived with the Trichopoulou model (2005). Participants were divided into three groups of adherence to Mediterranean diet (low, middle, and high tertiles). A cumulative midlife score was further derived for each dietary pattern. Confounding variables were father's social class, childhood cognition, educational attainment, adult social class, affective symptoms, smoking, and physical activity. FindingsBoth dietary patterns showed significant associations with memory decline in fully adjusted models. High quality diet predicted slower verbal memory decline compared with lower quality dietary choice; highest adherence to Mediterranean diet was similarly associated with slower memory decline compared with lowest adherence. These associations remained significant after adjustment for all covariates (regression coefficient 0·70, 95% CI 0·02–1·38, for the healthy dietary scale and 1·48, 0·17–2·80, for the highest adherence to Mediterranean diet). There were no associations at the 5% level between dietary choices and visual search decline. InterpretationOur results suggest that a healthy dietary choice, measured with either Mediterranean diet or healthy dietary scale in early midlife is protective of verbal memory decline from midlife to later life. Public health interventions based on healthy diets should be regarded as an important line of defence against cognitive decline and dementia. FundingDC, WN, AS, DK, and MR are funded by the UK Medical Research Council. HP is is supported by the Higher Education Funding Council for England. GM receives support from the Australian National Health and Medical Research Council (APP1000986).