We found that during IFNB treatment, increasing serum concentrations of alpha-tocopherol were associated with reduced odds for simultaneous and subsequent MRI disease activity. The results were not noticeably influenced by gender, age, BMI, HLA-DRB1*15 status, treatment group or compliance (omega-3 fatty acids or placebo), or the concentrations of NAb against IFNB, 25-hydroxyvitamin D, retinol, EPA or DHA.
Studies regarding vitamin E and MS are relatively few. In small cross-sectional studies, lower concentrations of both vitamin E and vitamin E/cholesterol-ratio have been reported in stable MS patients compared to controls 
, and also in MS patients during exacerbation compared to stable MS patients with or without IFNB treatment, and controls 
. In a prospective study, the risk of developing MS was not associated with total or dietary intake of vitamin E 
. To our knowledge, there are no prospective studies addressing the relationship between vitamin E and MRI disease activity in MS.
Our patient cohort comprises well characterized RRMS patients examined with repeated MRI scans and serum measurements, allowing eight paired MRI/alpha tocopherol assessments, and is well suited for a prospective study of the relationship between vitamin E and MS disease activity. Moreover, simultaneous measurements of vitamin A, vitamin D, DHA, EPA and NAb against IFNB in the same patients, combined with records of the compliance of study medication, enabled us to adjust for potential confounding. However, our study also has limitations. Although several paired MRI scans and serum measurements of alpha-tocopherol were conducted in each patient, the cohort might have been too small to detect minor, but nevertheless potentially important associations with relapse rate and EDSS progression. The dietary habits including use of vitamin supplements were not recorded. Moreover, the patients received either approximately 13 or 22 mg alpha-tocopherol from the omega-3 or placebo preparations. This constitutes 2–3 times the intake of total vitamin E estimated in a Finnish study 
, and also the estimated average daily intake of vitamin E in Norway (8–10 alpha-tocopherol equivalents, corresponding to 8–10 mg alpha-tocopherol) 
. Only the baseline values are therefore representative for the habitual vitamin E status of the patients. Accordingly, there was an increase in the mean concentration of alpha-tocopherol from baseline to the rest of the study period, and the supplementation might therefore have evened out both the inter- and intra-individual variation. Even so, we found a mean ratio between the highest and the lowest concentration of alpha-tocopherol in each patient of 1.34 and 22.8% of the total variation was accounted for by intra-individual variation. The serum concentration of alpha-tocopherol is correlated to the concentration of lipoproteins 
. Unfortunately, we did not measure lipoproteins or total cholesterol, and can therefore not exclude that lipoproteins might have confounded our results. However, in a previous study T2 lesion volume was not associated with lipoprotein concentrations 
, and the association with new T2 lesions was not influenced by BMI or omega-3 fatty acids in our study. It is therefore less likely that lipoprotein status have confounded the results.
Vitamin E has both antioxidative, immunomodulatory and neuroprotective properties 
. Our results are therefore biologically plauisible. Immune cells produce ROS, which may contribute to neuroinflammation in experimental allergic encephalomyelitis 
and MS 
. Vitamin E has also been shown to have other properties including regulation of enzymatic activity and gene transcription 
that might be relevant in MS 
, and to have immunomodulatory properties in animal models of rheumatoid arthritis 
and systemic lupus erythematosus 
Previous small studies have reported an increase of alpha-tocopherol in erythrocytes and plasma of MS patients treated with IFNB 
, as well as normalisation of ROS production in mononuclear cells 
. The finding that the odds for new MRI disease activity were only significantly reduced during IFNB treatment could indicate an interaction between IFNB and vitamin E. However, the difference in odds reduction before and during IFNB treatment was modest, and adjusting for NAb against IFNB did not alter our results. Thus, there is not sufficient evidence to draw any conclusion, and a possible interaction between vitamin E and IFNB treatment should be studied in a larger cohort.
In conclusion, we have shown an association between increasing alpha-tocopherol concentrations and simultaneous and subsequent MRI disease activity in RRMS patients during treatment with IFNB. The relation between vitamin E and MS should be further investigated in epidemiological and experimental studies.