Vitamin D deficiency is a known risk factor for multiple sclerosis (MS) and is associated with worsening disease activity and disability.
To examine the association between vitamin D with clinical outcomes, patient reported outcomes and quantitative MRI measures in a real-world MS cohort.
We conducted a retrospective analysis of patients enrolled in the Cleveland Clinic Clinical Practice Data Registry between June 2015 and November 2019. Serum 25-hydroxyvitamin D3 levels collected within 90 days of the first multiple sclerosis performance test (MSPT) assessment were recorded. Patients were dichotomized as sufficient or insufficient using a cutoff of 30 ng/mL. Baseline demographics, vitamin D supplementation, clinical outcome measures [Processing Speed Test (PST), Manual Dexterity Test (MDT), and Walking Speed Test (WST)], and patient reported outcome measures (PROMs) were collected. Brain/cervical MRIs obtained +/-90-days from the initial MSPT were analyzed via fully-automated methods for spinal cord cross sectional area (SCA), whole brain fraction (WBF), and T2 lesion volume (T2LV). Vitamin D associations were determined using Pearson correlation, quantile regression, and linear regression analysis.
369 patients (median age 47.3 years, 71.0% female, and 79.7% caucasian) were included in the analysis. Median age at diagnosis was 35 (IQR[interquartile range] 29.0-43.0), and median years with MS was 12.3 (IQR 5.49 – 20.6). Median vitamin D levels were 34.1 ng/mL [IQR 24.4;46.7] and 68.6% of patients were on supplementation. 62.3% (n=230) had sufficient vitamin D levels and 37.7% (n=139) had insufficient levels. No statistically significant differences were found between the groups for season at assessment, patient reported relapses, PST, MDT, or any quantitative MRI metrics The vitamin D insufficient [OD1] group had a statistically significantly longer WST (median 7.35 vs 6.56, p-value = 0.028). MDT dominant hand time had a statistically significant inverse relationship with vitamin D levels (ρ = -0.145) (p-value=0.005), which became non-significant after regression adjustment.
Vitamin D levels were found to be sufficient in two-thirds of patients and is likely explained by supplementation. Patients at sufficiency had faster walking speeds, but no other differences were found on clinical/MRI measures. We hypothesize that supplementation obscures the relation between vitamin D levels and clinical/MRI measures.