Objective: To investigate the association between 25-hydroxy vitamin D (25OHD) and cognitive function with particular consideration of physical activity (PA) in Icelandic older adults. Design: Cross-sectional study. Setting: Iceland. Participants: Old adults aged 65-96. The final analytical sample included 4304 non-demented participants. Measurements: Serum 25OHD was categorized into deficient (≤ 30 nmol/L, 8%), insufficient (31-49 nmol/L, 25%) and normal-high levels (>50 nmol/L, 67%). Cognitive function assessments included measurements of memory function (MF), speed of processing (SP) and executive function (EF) all categorized as low and high (divided by 50th percentile). Multivariate logistic regression analysis was used to calculate the odds ratio (OR) for having high cognitive function. Results: Serum 25OHD was positively associated with cognitive function. Adjustment for PA and other potential confounders diminished this association only partially. Compared to participants with normal-high levels of 25OHD, those with deficient levels had decreased odds for high SP (OR: 0.74, CI: 0.57-0.97), high MF (OR: 0.55; CI: 0.43-0.71) and high EF (OR: 0.76, CI: 0.57-1.0). Conclusion: Serum 25OHD below ≤30 nmol/L was associated with decreased odds for high cognitive function among community dwelling old adults as compared to those with 25OHD above > 50 nmol/L. Neither PA nor other potential confounders explained the associations between 25OHD and cognitive function. Future studies should explore mechanisms and the potential clinical relevance of this relationship.
All-cause mortality effects of replacing sedentary time with physical activity and sleeping using an isotemporal substitution model: a prospective study of 201,129 mid-aged and older adults