Vitamin D and Cognitive Performance in Older Adults: A Cross-Sectional and Mendelian Randomization Study

Cognitive decline is a prevalent health problem in older adults, and effective treatments remain to be produced. Serum vitamin D, a commonly used biochemical marker, is widely recognized as an indicator of various diseases. Existing research has not fully elucidated the relationship between vitamin D and cognitive function. The aim of this study is to investigate the real relationship between vitamin D and cognitive function and to identify indicators that have a strong predictive effect on cognitive decline. At first, we used the dataset of the genome-wide association studies studying vitamin D and cognitive performance to conduct Mendelian randomization analysis. Subsequently, we employed linear regression and smooth curve fitting methods to assess the relationship using the National Health and Nutrition Examination Survey data. Finally, we investigated other predictive features of cognitive performance utilizing a machine learning model. We found that a 1-unit increase in vitamin D is associated with a 6.51% reduction (P < .001) in the risk of cognitive decline. The correlation between vitamin D and cognitive performance is nonlinear, with the inflection point at 79.9 nmol/L (left: β = 0.043, P < .001; right: β = -0.007, P = .420). In machine learning, the top 5 predictors are vitamin D, weight, height, age, and body mass index. There is a causal relationship between vitamin D and cognitive performance. 79.9 nmol/L could be the optimal dose for vitamin D supplementation in the elderly. Further consideration of other factors in vitamin D interventions is necessary.