Abstract
25-Hydroxyvitamin D (25(OH)D) insufficiency is very common in many countries. Yet, the extent to which 25(OH)D status affects cognitive performance remains unclear. The objective of the present study was to evaluate the cross-time association between midlife plasma 25(OH)D concentrations and subsequent cognitive performance, using a subsample from the French 'SUpplémentation en Vitamines et Minéraux AntioXydants' randomised trial (SU.VI.MAX, 1994-2002) and the SU.VI.MAX 2 observational follow-up study (2007-9). 25(OH)D concentrations were measured in plasma samples drawn in 1994-5, using an electrochemoluminescent immunoassay. Cognitive performance was evaluated in 2007-9 with a neuropsychological battery including phonemic and semantic fluency tasks, the RI-48 (rappel indicé-48 items) cued recall test, the Trail Making Test and the forward and backward digit span. Cognitive factors were extracted via principal component analysis (PCA). Data from 1009 individuals, aged 45-60 years at baseline, with available 25(OH)D and cognitive measurements were analysed by multivariable linear regression models and ANCOVA, stratified by educational level. PCA yielded two factors, designated as 'verbal memory' (strongly correlated with the RI-48 and phonemic/semantic fluency tasks) and 'short-term/working memory' (strongly correlated with the digit span tasks). In the fully adjusted regression model, among individuals with low education, there was a positive association between 25(OH)D concentrations and the 'short-term/working memory' factor (P=0.02), mainly driven by the backward digit span (P=0.004). No association with either cognitive factor was found among better educated participants. In conclusion, higher midlife 25(OH)D concentrations were linked to better outcomes concerning short-term and working memory. However, these results were specific to subjects with low education, suggesting a modifying effect of cognitive reserve.
