Objective: We aimed to test the dose-response association of serum 25(OH)D and risk of dementia and Alzheimer's disease (AD). Methods: We performed a systematic search of PubMed and Scopus from database inception up to September 2017. Longitudinal cohort studies reporting risk estimates of incident dementia or AD in the general population, and for three or more quantitative categories of serum 25(OH)D were included. Pooled hazard ratios (HRs) were calculated using fixed-effects/random-effects models. Results: Seven prospective cohort studies and one retrospective cohort study (total n = 28,354) involving 1953 cases of dementia and 1607 cases of AD were included. The pooled HRs of dementia and AD were 1.09 (95%CI: 0.95, 1.24) and 1.19 (95%CI: 0.96, 1.41) for vitamin D insufficiency (10-20 ng/ml), and 1.33 (95%CI: 1.08, 1.58) and 1.31 (95%CI: 0.98, 1.65) for deficiency (<10 ng/ml), respectively. The lower risk of dementia was observed at serum 25(OH)D of ∼25 ng/ml, whereas the risk of AD decreased continuously along with the increase of serum 25(OH)D up to ∼35 ng/ml. Conclusion: Higher levels of serum 25(OH)D was associated with a lower risk of dementia and AD, but we have no conclusive evidence regarding serum 25(OH)D levels of >35 ng/ml.
Keywords: Alzheimer disease; Dementia; Meta-analysis; Vitamin D deficiency.