Abstract
OBJECTIVES: Research concerning the link between dairy product intake and cognition is scant while experimental studies suggest links through various biological mechanisms. This study's objective was to examine the cross-time associations of total and specific dairy product consumption with cognitive performance in aging adults. We also explored compliance with dairy intake recommendations in France. DESIGN: The study was based on the «Supplémentation en Vitamines et Minéraux Antioxydants» randomized trial (SU.VI.MAX, 1994-2002) and the SU.VI.MAX 2 observational follow-up study (2007-2009). SETTING: A general-population cohort in France. PARTICIPANTS: N=3,076 participants included in both the SU.VI.MAX and SU.VI.MAX 2 studies. MEASUREMENTS: Dairy product consumption was estimated using repeated 24h records (1994-1996; mean=10 records, SD=3). Cognitive performance was assessed by neuropsychologists after an average of 13 years post-baseline via a battery of six validated tests. Mean age at the time of the cognitive function evaluation was 65.5 (SD=4.6) years. Principal component analysis revealed factors for verbal memory and working memory. Associations of energy-adjusted dairy product consumption and compliance with the respective dietary guidelines with subsequent cognitive impairment were examined using ANCOVA, providing mean differences (95% confidence intervals, CI) according to tertiles (T), adjusted for confounders including overall dietary patterns. RESULTS: Total dairy product consumption was not associated with cognitive function. However, milk intake was negatively associated with verbal memory performance: mean difference T3 versus T1= -0.99 (-1.83, -0.15). Among women, consuming more than the recommended amount of dairy was negatively associated with working memory performance: excess versus adequate = -1.52 (-2.93, -0.11). CONCLUSION: Our results indicate that dairy products consumption and especially compliance with dietary guidelines regarding dairy product intake are differentially associated with performance in specific cognitive domains after a comprehensive adjustment for lifestyle factors, health status markers and dietary patterns. Further longitudinal research is needed given the limited data available.