Vitamin D and the Risk of Depression: A Causal Relationship? Findings from a Mendelian Randomization Study

Nutrients
Q1
May 2019
Citations:47
Influential Citations:1
Observational Studies (Human)
81
S2 IconPDF Icon

Enhanced Details

Methods
Two-sample Mendelian randomization using GWAS summary statistics. Exposure: serum 25-hydroxyvitamin D levels from SUNLIGHT (European ancestry; n=79,366). Outcomes: depressive symptoms (DS) (n=161,460) and broad depression (BD) (113,769 cases, 208,811 controls) from UK Biobank. Instrument: six genome-wide significant SNPs explaining 2.8% of vitamin D variance. Study design: IVW and MR-Egger analyses with sensitivity checks (leave-one-out). Instrument strength indicated by F-statistic (381).
Results
No evidence that genetically higher 25(OH) vitamin D levels causally reduce risk of depressive symptoms or broad depression. IVW p-values: DS 0.52; BD 0.10. MR-Egger intercepts showed no pleiotropy. Leave-one-out analyses were consistent. Implication: raising vitamin D levels is unlikely to prevent depression in European populations; observational associations may reflect confounding or reverse causation. Power was 100% to detect an odds ratio of 1.1 per 1 SD decrease in ln(25(OH)D) for BD; smaller effects cannot be ruled out.
Limitations
SNPs explained 2.8% of vitamin D variance (weak instrument concern) though F-statistic = 381 indicates strong instruments; analyses limited to individuals of European ancestry; possible horizontal pleiotropy cannot be fully ruled out (MR-Egger intercept non-significant); power limits: able to detect only moderate effects (OR ≥1.1); DS and BD are broad depression phenotypes; results may not generalize to subtypes or other populations.

Abstract

While observational studies show an association between 25(OH)vitamin D concentrations and depressive symptoms, intervention studies, which examine the preventive effects of vitamin D supplementation on the development of depression, are lacking. To ...