To evaluate the impact of plasma vitamin D (VD) increase on the 24-months risk of disease activity (DA, defined as a relapse and/or MRI activity (new and/or contrast-enhancing lesions)).

VD levels were measured centrally by mass spectrometry at baseline and 3 months to reflect the overall change in VD during the study. Groups were compared using a Wilcoxon test. Spearman R correlation coefficient was used for correlations. The effect of VD on DA was analysed using ROC curves and adjusted Cox models.

266 patients were analyzed (n=140 in the VD group and n=126 in the placebo group) and showed similar median VD levels (52.0 vs. 49.3 nmol/L). In both groups, VD levels at baseline did not predict DA. VD increase at 3 months discriminated DA vs. no DA in the placebo group (median increase 1.6 nmol/L, p=0.036) but not in the VD group (median increase 94.1 nmol/L, p=0.602), where it weakly correlated to patients’ weight (R=-0.22, p=0.031) and to body mass index (BMI, R=-0.37, p<0.001). Combining both groups, VD increase significantly reduced the risk of DA (HR=0.58 for 100 nmol/L increase, p=0.008).

In line with the global results of D-lay MS, VD increase after supplementation with 100,000 IU cholecalciferol every two weeks significantly reduced the risk of DA at the cohort level. Interestingly, variations of VD levels at 3 months were predictive of DA in the placebo group but not in the VD group, suggesting a maximal effect in this high dosage. Overall, given the excellent safety of VD, these findings suggest that individual adjustment of this high-dose supplementation regimen may be unnecessary.