[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 54.87.121.0. Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Download PDF
Figure 1.
Relative Rate of Cumulative New Active Lesions (NALs) vs Average of Baseline, 6-Month, and 12-Month 25-Hydroxyvitamin D (25[OH]D) Level Stratified by Geographic Region
Relative Rate of Cumulative New Active Lesions (NALs) vs Average of Baseline, 6-Month, and 12-Month 25-Hydroxyvitamin D (25[OH]D) Level Stratified by Geographic Region

The solid lines and shaded regions represent the relative rate ratios of cumulative NALs for changes in 25(OH)D relative to the median level and the corresponding 95% CIs, respectively. Analyses were adjusted for age, sex, randomization status, baseline Expanded Disability Status Scale score, and disease duration (≤1, 2-5, 6-10, and >10 years). Models assume a linear association between the logarithm of the rate of cumulative NALs and serum 25(OH)D. Analyses using cubic splines revealed no significant deviation from linearity. To convert 25(OH)D values to nanograms per milliliter, divide by 2.496.

Figure 2.
Relative Rate of Cumulative New Active Lesions (NALs) vs Average of Baseline, 6-Month, and 12-Month 25-Hydroxyvitamin D Level Stratified by Baseline Disease Characteristics
Relative Rate of Cumulative New Active Lesions (NALs) vs Average of Baseline, 6-Month, and 12-Month 25-Hydroxyvitamin D Level Stratified by Baseline Disease Characteristics

From 1482 patients eligible for the analysis, multivariable models excluded patients with missing follow-up information on subsequent magnetic resonance imaging (MRI) (n = 26), patients with missing information for baseline Expanded Disability Status Scale (EDSS) score (n = 2), and patients with missing disease duration (n = 1), leaving 1453 participants in the final models. High disease activity was defined as having at least 2 relapses in the previous 2 years (to study baseline) and at least 1 T1-enhancing lesion on baseline MRI.

Table 1.  
Selected Age- and Region-Adjusted Characteristics of BEYOND Participantsa
Selected Age- and Region-Adjusted Characteristics of BEYOND Participantsa
Table 2.  
MRI Evidence of Disease Activity per 50.0-nmol/L Increase in 25(OH)D in Cross-Sectional Analyses
MRI Evidence of Disease Activity per 50.0-nmol/L Increase in 25(OH)D in Cross-Sectional Analyses
Table 3.  
Categorical Season-Adjusted 25(OH)D and Rate of New Active Lesions
Categorical Season-Adjusted 25(OH)D and Rate of New Active Lesions
1.
Rudick  RA, Miller  D, Hass  S,  et al; AFFIRM and SENTINEL Investigators.  Health-related quality of life in multiple sclerosis: effects of natalizumab. Ann Neurol. 2007;62(4):335-346.
PubMedArticle
2.
Frohman  EM, Racke  MK, Raine  CS.  Multiple sclerosis—the plaque and its pathogenesis. N Engl J Med. 2006;354(9):942-955.
PubMedArticle
3.
Munger  KL, Zhang  SM, O’Reilly  E,  et al.  Vitamin D intake and incidence of multiple sclerosis. Neurology. 2004;62(1):60-65.
PubMedArticle
4.
Munger  KL, Levin  LI, Hollis  BW, Howard  NS, Ascherio  A.  Serum 25-hydroxyvitamin D levels and risk of multiple sclerosis. JAMA. 2006;296(23):2832-2838.
PubMedArticle
5.
Banwell  B, Bar-Or  A, Arnold  DL,  et al.  Clinical, environmental, and genetic determinants of multiple sclerosis in children with acute demyelination: a prospective national cohort study. Lancet Neurol. 2011;10(5):436-445.
PubMedArticle
6.
Salzer  J, Hallmans  G, Nyström  M, Stenlund  H, Wadell  G, Sundström  P.  Vitamin D as a protective factor in multiple sclerosis. Neurology. 2012;79(21):2140-2145.
PubMedArticle
7.
Ascherio  A, Munger  KL, White  R,  et al.  Vitamin D as an early predictor of multiple sclerosis activity and progression. JAMA Neurol. 2014;71(3):306-314.
PubMedArticle
8.
Martinelli  V, Dalla Costa  G, Colombo  B,  et al.  Vitamin D levels and risk of multiple sclerosis in patients with clinically isolated syndromes. Mult Scler. 2014;20(2):147-155.
PubMedArticle
9.
Simpson  S  Jr, Taylor  B, Blizzard  L,  et al.  Higher 25-hydroxyvitamin D is associated with lower relapse risk in multiple sclerosis. Ann Neurol. 2010;68(2):193-203.
PubMed
10.
Runia  TF, Hop  WC, de Rijke  YB, Buljevac  D, Hintzen  RQ.  Lower serum vitamin D levels are associated with a higher relapse risk in multiple sclerosis. Neurology. 2012;79(3):261-266.
PubMedArticle
11.
Løken-Amsrud  KI, Holmøy  T, Bakke  SJ,  et al.  Vitamin D and disease activity in multiple sclerosis before and during interferon-β treatment. Neurology. 2012;79(3):267-273.
PubMedArticle
12.
Mowry  EM, Krupp  LB, Milazzo  M,  et al.  Vitamin D status is associated with relapse rate in pediatric-onset multiple sclerosis. Ann Neurol. 2010;67(5):618-624.
PubMed
13.
Devonshire  V, Havrdova  E, Radue  EW,  et al; FREEDOMS Study Group.  Relapse and disability outcomes in patients with multiple sclerosis treated with fingolimod: subgroup analyses of the double-blind, randomised, placebo-controlled FREEDOMS study. Lancet Neurol. 2012;11(5):420-428.
PubMedArticle
14.
O’Connor  P, Filippi  M, Arnason  B,  et al; BEYOND Study Group.  250 µg or 500 µg Interferon beta-1b versus 20 mg glatiramer acetate in relapsing-remitting multiple sclerosis: a prospective, randomised, multicentre study. Lancet Neurol. 2009;8(10):889-897.
PubMedArticle
15.
Lampl  C, Nagl  S, Arnason  B,  et al.  Efficacy and safety of interferon beta-1b sc in older RRMS patients—a post hoc analysis of the BEYOND study. J Neurol. 2013;260(7):1838-1845.
PubMedArticle
16.
Smith  SM, Zhang  Y, Jenkinson  M,  et al.  Accurate, robust, and automated longitudinal and cross-sectional brain change analysis. Neuroimage. 2002;17(1):479-489.
PubMedArticle
17.
Li  Y, Willer  CJ, Ding  J, Scheet  P, Abecasis  GR.  MaCH: using sequence and genotype data to estimate haplotypes and unobserved genotypes. Genet Epidemiol. 2010;34(8):816-834.
PubMedArticle
18.
Wang  YC, Meyerson  L, Tang  YQ, Qian  N.  Statistical methods for the analysis of relapse data in MS clinical trials. J Neurol Sci. 2009;285(1-2):206-211.
PubMedArticle
19.
Andersen  PK, Gill  RD.  Cox’s regression model counting process: a large sample study. Ann Stat. 1982;10:1100-1120.Article
20.
Mowry  EM, Waubant  E, McCulloch  CE,  et al.  Vitamin D status predicts new brain magnetic resonance imaging activity in multiple sclerosis. Ann Neurol. 2012;72(2):234-240.
PubMedArticle
21.
ClinicalTrials.gov. Supplementation of VigantOL Oil Versus Placebo as Add-on in Patients With Relapsing Remitting Multiple Sclerosis Receiving Rebif Treatment (SOLAR). NCT01285401. https://clinicaltrials.gov/ct2/show/NCT01285401. Accessed August 31, 2015.
22.
Becklund  BR, Severson  KS, Vang  SV, DeLuca  HF.  UV radiation suppresses experimental autoimmune encephalomyelitis independent of vitamin D production. Proc Natl Acad Sci U S A. 2010;107(14):6418-6423.
PubMedArticle
23.
Bertrand  KA, Giovannucci  E, Liu  Y,  et al.  Determinants of plasma 25-hydroxyvitamin D and development of prediction models in three US cohorts. Br J Nutr. 2012;108(10):1889-1896.
PubMedArticle
24.
Bock  G, Prietl  B, Mader  JK,  et al.  The effect of vitamin D supplementation on peripheral regulatory T cells and β cell function in healthy humans: a randomized controlled trial. Diabetes Metab Res Rev. 2011;27(8):942-945.
PubMedArticle
25.
Mahon  BD, Gordon  SA, Cruz  J, Cosman  F, Cantorna  MT.  Cytokine profile in patients with multiple sclerosis following vitamin D supplementation. J Neuroimmunol. 2003;134(1-2):128-132.
PubMedArticle
26.
Burton  JM, Kimball  S, Vieth  R,  et al.  A phase I/II dose-escalation trial of vitamin D3 and calcium in multiple sclerosis. Neurology. 2010;74(23):1852-1859.
PubMedArticle
27.
Smolders  J, Peelen  E, Thewissen  M,  et al.  Safety and T cell modulating effects of high dose vitamin D3 supplementation in multiple sclerosis. PLoS One. 2010;5(12):e15235. doi:10.1371/journal.pone.0015235.
PubMedArticle
28.
Jorde  R, Sneve  M, Torjesen  PA, Figenschau  Y, Gøransson  LG, Omdal  R.  No effect of supplementation with cholecalciferol on cytokines and markers of inflammation in overweight and obese subjects. Cytokine. 2010;50(2):175-180.
PubMedArticle
29.
Schleithoff  SS, Zittermann  A, Tenderich  G, Berthold  HK, Stehle  P, Koerfer  R.  Vitamin D supplementation improves cytokine profiles in patients with congestive heart failure: a double-blind, randomized, placebo-controlled trial. Am J Clin Nutr. 2006;83(4):754-759.
PubMed
Original Investigation
December 2015

Association of Vitamin D Levels With Multiple Sclerosis Activity and Progression in Patients Receiving Interferon Beta-1b

Author Affiliations
  • 1Department of Nutrition, Harvard T. H. Chan School of Public Health, Harvard University, Boston, Massachusetts
  • 2Bayer Pharma AG, Berlin, Germany
  • 3Department of Neurology, University of Chicago, Chicago, Illinois
  • 4Department of Neurology, Scientific Institute H. S. Raffaele, Milan, Italy
  • 5Department of Neurology and Neuroscience, Rutgers–New Jersey Medical School, Newark
  • 6Department of Neurology, University of California, San Francisco
  • 7Neuroimaging Research Unit, Scientific Institute and University Vita-Salute, Milan, Italy
  • 8Department of Neurology, University of Dusseldorf, Dusseldorf, Germany
  • 9Department of Neurology, Piedmont Health Care, Mooresville, North Carolina
  • 10Department of Neurology, St Michael’s Hospital, Toronto, Ontario, Canada
  • 11Bayer HealthCare Pharmaceuticals, Whippany, New Jersey
  • 12Bayer Pharma AG, Berlin, Germany
  • 13Department of Neurology, University Hospital Basel, Basel, Switzerland
  • 14Department of Neurology, University Hospital Bonn, Bonn, Germany
  • 15Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts
  • Christoph Pohl, MD
JAMA Neurol. 2015;72(12):1458-1465. doi:10.1001/jamaneurol.2015.2742
Abstract

Importance  Low serum 25-hydroxyvitamin D (25[OH]D) levels are associated with an increased risk of multiple sclerosis (MS) as well as with increased disease activity and rate of progression in clinically isolated syndromes and early MS.

Objective  To assess the association between 25(OH)D and disease course and prognosis in patients with relapsing-remitting MS treated with interferon beta-1b.

Design, Setting, and Participants  We conducted a prospective cohort study assessing 25(OH)D levels and subsequent MS disease course and progression characterized by magnetic resonance imaging (MRI) and clinical end points. The study took place between November 2003 and June 2005; data analysis was performed between June 2013 and December 2014. The study was conducted among participants in the Betaferon Efficacy Yielding Outcomes of a New Dose (BEYOND) study, a large, phase 3, prospective, multicenter, blinded, randomized clinical trial. Patients were monitored for at least 2 years. Clinic visits were scheduled every 3 months, and MRI was performed at baseline and annually thereafter. Eligible patients included 1482 participants randomized to receive 250 μg or 500 μg of interferon-1b with at least 2 measurements of 25(OH)D obtained 6 months apart.

Exposures  Serum 25(OH)D measurements were performed at baseline, 6 months, and 12 months.

Main Outcomes and Measures  Main outcomes included cumulative number of new active lesions (T2 lesions and gadolinium acetate–enhancing lesions), change in normalized brain volume, relapse rate, and progression determined by the Expanded Disability Status Scale (EDSS). Statistical analyses were adjusted for age, sex, randomized treatment, region, disease duration, and baseline EDSS score.

Results  Overall, average 25(OH)D levels in 1482 patients were significantly inversely correlated with the cumulative number of new active lesions between baseline and the last MRI, with a 50.0-nmol/L increase in serum 25(OH)D levels associated with a 31% lower rate of new lesions (relative rate [RR], 0.69; 95% CI, 0.55-0.86; P = .001). The lowest rate of new lesions was observed among patients with 25(OH)D levels greater than 100.0 nmol/L (RR, 0.53; 95% CI, 0.37-0.78; P = .002). No significant associations were found between 25(OH)D levels and change in brain volume, relapse rates, or EDSS scores. Results were consistent following adjustment for HLA-DRB1*15 or vitamin D–binding protein status.

Conclusions and Relevance  Among patients with MS treated with interferon beta-1b, higher 25(OH)D levels were associated with lower rates of MS activity observed on MRI. Results for brain atrophy and clinical progression were more equivocal.

Introduction

Multiple sclerosis (MS) is a debilitating disease of the central nervous system characterized by episodic periods of inflammatory demyelination and progressive neurologic deficits.1,2 The higher MS risk among individuals with low vitamin D intake3 or low circulating 25-hydroxyvitamin D (25[OH]D)46 levels and the worse prognosis among patients with MS who have low 25(OH)D levels712 suggest that these low levels may contribute to the disease process. Most recently, a previous study indicated that individuals with average serum 25(OH)D levels less than <50.0 nmol/L (to convert to nanograms per milliliter, divide by 2.496) in the 12 months following a first demyelinating event had worse clinical and imaging outcomes after 5 years than did those with higher levels.7 However, since average 25(OH)D levels in that study were low and there was no clear ceiling effect, the optimal serum level of 25(OH)D could not be identified. In addition, it is important to determine whether this association between 25(OH)D and MS outcomes differs among subgroups of patients and particularly whether it applies to patients with high disease activity who may benefit from other potential therapeutic alternatives.13 To address these issues, we conducted a prospective study among a large group of patients with relapsing-remitting MS treated with interferon beta-1b with a wide distribution of serum 25(OH)D levels and disease severity.

Methods
Study Population

Eligible patients included 1796 individuals with MS randomized to receive either 250 μg or 500 μg of interferon beta-1b (Betaferon) in the Betaferon Efficacy Yielding Outcomes of a New Dose (BEYOND) trial.14 The present study was approved by the Harvard T. H. Chan School of Public Health. Participants in the BEYOND trial provided written informed consent. The present study used de-identified data and biological samples and had no direct contact with BEYOND participants. Full descriptions of the study population and results have been reported.14,15 Briefly, BEYOND was a large, phase 3, prospective, multicenter, blinded, randomized clinical trial with interferon beta-1b, 250 µg (n = 897); interferon beta-1b, 500 µg (n = 899); and glatiramer acetate, 20 mg (n = 448), treatment arms. Following enrollment from November 2003 to June 2005, patients were monitored for at least 2 years. Clinic visits were scheduled every 3 months for assessment of relapse, Expanded Disability Status Scale (EDSS) score, safety, and tolerability, and magnetic resonance imaging (MRI) was performed at baseline and annually thereafter. Both T2- and T1-weighted images (with and without 0.1 mmoL/kg of gadolinium-diethylenetriamine pentaacetic acid) of the whole brain were obtained and evaluated in a blinded manner by the Neuroimaging Research Unit in Milan, Italy. Observers identified T1- and T2-weighted lesions, and a semiautomated local threshold technique was used to determine lesion volume.16 Normalized brain volume at baseline was assessed using Sienax and longitudinal brain volume percent change was evaluated with Siena.16

Levels of 25(OH)D were measured in serum samples collected at baseline, 6 months, and 12 months from patients randomized to receive 250 μg or 500 μg of interferon beta-1b; patients randomized to glatiramer acetate were not included. Because the primary purpose of the study was to estimate the effects of yearly average 25(OH)D levels, we included in the analyses only the 1482 (from the 1796 eligible) patients with at least two 25(OH)D measurements performed 6 months apart; all of these patients had 25(OH)D levels determined at 6 months, 1465 patients (98.9%) had the baseline measurement, and 1426 individuals (96.2%) had 25(OH)D levels obtained at 12 months. Of these 1482 participants, 1456 (98.2%) had follow-up information on MRI measurements after the baseline assessment. Patients from 26 countries participated in the BEYOND trial14 and were initially grouped into 9 countries/regions as follows: Western Europe (Austria, Belgium, Switzerland, Germany, and the Netherlands), the Mediterranean (Spain, France, Italy, Greece, and Israel), Scandinavia (Denmark, Sweden, Finland, and Norway), Eastern Europe (Ukraine, Latvia, Hungary, Slovenia, and Poland), North America (United States and Canada), Russia, Argentina, Brazil, and Australia. For the purpose of adjusting the primary analyses of 25(OH)D levels and MS outcomes, these regions were further collapsed into 4 broad areas based both on geography and patient characteristics: Eastern Europe (including Russia), Western Europe (including Israel), North America, and the southern hemisphere (Australia, Argentina, and Brazil). Although these categories are arbitrary and not entirely homogeneous (eg, 25[OH]D levels were lower in Russia than Eastern Europe), sensitivity analyses, in which confounding by geography was controlled using the original 9 regions, gave virtually identical results.

Measurement of 25(OH)D

Samples were shipped on dry ice to the central laboratory in Schwentinental, Germany, within 3 days of being drawn and then maintained at −20°C until further analysis. Serum 25(OH)D was measured separately in each treatment group (250 or 500 μg of interferon beta-1b) using a chemiluminescence assay (Roche Diagnostics) in June 2013. The average intra-assay coefficients of variation derived from blind quality control samples were 2.6% for the samples from patients randomized to 250 μg of interferon beta-1b and 4.0% for patients randomized to 500 μg of interferon beta-1b. The corresponding between-assay coefficients of variation were 5.0% and 7.1%, and the intraclass correlation for quality control samples run with both the 250- and the 500-μg doses of interferon beta-1b was 0.95. Baseline level of 25(OH)D was strongly correlated with levels at both 6 months (Spearman correlation coefficient, r = 0.75) and 12 months (r = 0.89); 25(OH)D levels displayed the expected seasonal variations. Because we were interested in the effects of long-term average 25(OH)D levels and as a result of large between-region variation, we performed season adjustment as previously described4 within each region (using the 4-region categorization).

Genotype Assessment

All samples were genotyped (GeneChip, version 6.0 platform; Affymetrix). Both HLA-DRB1*15 and vitamin D–binding protein were imputed with MaCH17 using the 120 phased chromosomes from the CEU (Utah residents with ancestry from northern and western Europe) or samples in HapMap phase II (http://hapmap.ncbi.nlm.nih.gov/) as the reference.

Statistical Analysis

Our data analysis was performed from June 2013 to December 2014. We treated the season-adjusted serum 25(OH)D level as a time-dependent variable and considered the value at each time point to be an average of all previous values. The association between season-adjusted 25(OH)D levels and MS outcomes was examined in both cross-sectional and longitudinal analyses. In the cross-sectional analysis, we assessed the association between baseline season-adjusted serum 25(OH)D levels with baseline MRI measurements (including volume of T2 hyperintense lesions, number of T1-enhancing lesions, volume of T1-nonenhancing hypointense lesions, and normalized brain volume) and clinical measurements (including baseline EDSS score and number of relapses in the 2 years before study baseline). We also assessed the cross-sectional association between the cumulative average mean 25(OH)D level at 12 months and MRI measurements performed at year 1. In longitudinal analyses, we considered the association of time-dependent, season-adjusted average 25(OH)D levels with the cumulative number of new active lesions (the sum of new T2 lesions and new gadolinium-enhancing lesions), change in normalized brain volume, relapse rate, and confirmed EDSS score progression (defined as an increase of 1 point compared with baseline and sustained for ≥3 months).

We modeled lesion counts using negative binomial regression and continuous outcomes (T2 lesion volume and normalized brain volume) using generalized linear regression (applying log transformations to skewed data), allowing for patient-specific random intercepts in longitudinal analyses.18 We assessed relapse rate using an Andersen-Gill model19 for recurrent events and assessed time to confirmed EDSS score progression using Cox regression analysis.

Because of large differences in 25(OH)D levels across regions, categorization using either quintiles or prespecified cut points of 25(OH)D levels resulted in groups that were largely region dependent. Given these differences and the varied baseline disease characteristics across regions, in primary analyses, we modeled baseline and cumulative averaged 25(OH)D using linear terms. We assessed the robustness of this modeling strategy using restricted cubic splines and found no evidence of thresholds or departure from linearity in any model for either MRI or clinical outcome. Further, we conducted analyses restricted to individuals with average 25(OH)D levels of at least 50.0 nmol/L and, separately, categorical analyses using the following a priori categories of average 25(OH)D values: less than 37.5, 37.5-49.9, 50.0-74.9, 75.0-99.9, and 100.0 nmol/L or higher (both overall and stratified by region).

All analyses were adjusted for age, sex, randomized treatment (interferon beta-1b, 250 or 500 μg), disease duration (≤1 year, 2-5 years, 6-10 years, and >10 years), and region (North America, Western Europe, Eastern Europe, and the southern hemisphere, with the exception of analyses stratified by region). Longitudinal analyses were additionally adjusted for baseline disease characteristics, including EDSS score, and, in sensitivity analyses, for baseline T2 lesion volume and race/ethnicity. We also assessed the effects of 25(OH)D following prespecified subgroups based on baseline characteristics: age (median or less vs greater than the median), sex, relapses in the previous 2 years (<2 vs ≥2), disease duration (≤1 year, 2-5 years, 6-10 years, and >10 years), T2-lesion volume (median or less vs greater than the median), EDSS score (<3 vs ≥3), and disease activity (high vs low). High disease activity was defined as having at least 2 relapses in the previous 2 years (to study baseline) and at least 1 T1-enhancing lesion noted on baseline MRI.

We also conducted analyses assessing the association between MS disease activity and serum 25(OH)D levels adjusting for HLA-DRB1 (using rs3135388 as a marker for DRB1*15; imputation r2 >0.99) or vitamin D–binding protein/GC (using rs4588 as a marker for GC-1s/f vs GC-2; imputation r2 = 0.96) and assessed whether either single-nucleotide polymorphism modified such association.

Results

Among the 1796 individuals with MS included in our study, average 25(OH)D levels varied markedly by region: the lowest levels were observed in Russia (median, 41.9 nmol/L), and the highest levels were documented in Australia (median, 66.4 nmol/L), Brazil (68.9 nmol/L), and North America (59.7 nmol/L) (eTable 1 in the Supplement). Significant differences were also observed in several clinical parameters. The most salient differences were the higher volume of T2 hyperintense lesions, younger age at onset, and longer disease duration among patients from Eastern Europe and Russia compared with those from Western Europe and North America (eTable 1 in the Supplement). Following standardization by age and region, at baseline patients with higher 25(OH)D levels tended to have lower body mass index, a lower EDSS score and Multiple Sclerosis Severity Score, fewer relapses in the previous 2 years, lower T1 and T2 lesion volumes, and higher normalized brain volume (Table 1).

MRI Evidence of Disease Activity

In cross-sectional analyses adjusted for age, sex, randomization treatment group (interferon beta-1b, 250 vs 500 μg), baseline EDSS score, disease duration, and region of residence, baseline 25(OH)D levels, reported as estimates (95% CIs), were inversely associated with T2 hyperintense lesion volume (a 50-nmol/L higher level corresponding to a difference of −0.11 cm3 in log[T2 lesion volume]; −0.20 to −0.02; P = .02) and the total number of T1-enhancing lesions (−0.24; −0.45 to −0.02; P = .03) at baseline (Table 2). Similar results were obtained for the association between 25(OH)D and MRI parameters at 12 months, with a difference in log(T2 lesion volume) of −0.19 cm3 (−0.31 to 0.07) (Table 2). Participant 25(OH)D levels were also inversely associated with the cumulative number of new active lesions (defined as the sum of new T2- and T1-enhancing lesions) between baseline and 12 months (0.76; 0.60 to 0.98; P = .03) (Table 2).

In longitudinal analyses, 25(OH)D was significantly inversely correlated with the cumulative number of active lesions between baseline and the last MRI (average follow-up time, 2 years), with a 50-nmol/L higher level in serum 25(OH)D levels associated with a 31% lower rate of new lesions (relative rate [RR], 0.69; 95% CI, 0.55-0.86; P = .001). This inverse association was also strong and significant in analyses restricted to patients with 25(OH)D levels greater than 50.0 nmol/L (RR, 0.62; 95% CI, 0.46-0.84; P = .002), and was consistently observed in each of the 4 regions (Figure 1). The dose-response association between 25(OH)D and the cumulative number of active lesions was further explored in categorical analyses stratified by region (Table 3). Patients with serum 25(OH)D levels greater than >100.0 nmol/L had a 47% lower rate of new active lesions compared with patients who had serum levels of 50.0 to 74.9 nmol/L (RR, 0.53; 95% CI, 0.37-0.78; P = .002) and were relatively consistent within the region (Table 3). Following the potential threshold effect we observed (Table 3), we fit an additional regression model to test for a threshold at 100.0 nmol/L. Results of this analysis suggested no significant differences in the effect of 25(OH)D on the rate of cumulative new active lesions for individuals with levels less than 100.0 nmol/L and those with levels of 100.0 nmol/L or more (P = .39). In analyses further adjusted for HLA-DRB1*15 or vitamin D–binding protein status, results were slightly stronger (adjustment for HLA-DRB1*15: RR per 50.0 nmol/L, 0.65; 95% CI, 0.46-0.93; P = .02; adjustment for vitamin D–binding protein: RR per 50.0 nmol/L, 0.63; 95% CI, 0.45-0.90; P = .01). We did not observe evidence of an interaction between single-nucleotide polymorphism or 25(OH)D levels and MS.

In additional analyses, the association between serum 25(OH)D and MS activity was slightly attenuated following adjustment for race/ethnicity and baseline T2 lesion volume but remained statistically significant (RR of cumulative new active lesions for 50.0-nmol/L increase, 0.74; 95% CI, 0.60-0.92; P = .009). Results were also consistent in analyses using the 12-month MRI measurements as the baseline; in these analyses, a 50.0-nmol/L higher level in serum 25(OH)D levels was associated with a 25% lower rate of new lesions (RR, 0.75; 95% CI, 0.60-0.94; P = .01). Similarly, additional analyses adjusting for body mass index did not materially influence the results (for cumulative number of active lesions: RR for 50.0-nmol/L increase, 0.69; 95% CI, 0.55-0.86; P = .001). The inverse association between average 25(OH)D levels and the cumulative number of new active lesions was similar across strata in exploratory analyses defined by age, sex, disease activity and severity at baseline, disease duration, T2 lesion volume, and EDSS score (Figure 2).

Relapses and EDSS

No significant association was found in a cross-sectional analysis between baseline 25(OH)D levels and the number of relapses in the previous 2 years (relapse rate, 1.01; 95% CI, 0.94 to 1.09; P = .77) or between 25(OH)D level and the number of relapses from baseline to the last visit (relapse rate, 1.01; 95% CI, 0.84 to 1.21; P = .92). Baseline 25(OH)D level was inversely associated with EDSS scores at baseline: a 50.0-nmol/L higher level of 25(OH)D corresponded to a mean difference of 0.28 in EDSS scores (mean difference, −0.28; 95% CI, −0.40 to −0.16; P < .001). However, analyses were attenuated in longitudinal analyses adjusted for age, sex, randomization status, region of residence, and baseline disease duration, with 25(OH)D no longer significantly associated with confirmed EDSS progression (defined as a 1-point increase from baseline that was sustained for ≥3 months; RR per 50.0-nmol/L increase: 0.84; 95% CI, 0.86 to 1.15; P = .14). Results were slightly stronger using month 12 as the baseline but remained insignificant (confirmed EDSS progression: RR per 50.0-nmol/L increase, 0.77; 95% CI, 0.56 to 1.06; P = .11).

Brain Volume

No associations were found between 25(OH)D levels and brain volume at baseline or at 12 months. Similarly, we did not observe an association between 25(OH)D or relative change in brain volume from baseline or 12 months to 24 months (eTable 2 in the Supplement).

Discussion

In this large prospective study among patients with relapsing-remitting MS treated with interferon beta-1b and with median disease duration of 3 years, we found that, in cross-sectional analyses, those with higher serum 25(OH)D levels had a lower disease burden noted on MRI and a lower baseline EDSS score. Furthermore, patients with higher serum 25(OH)D levels during the first 12 months of the study developed fewer new active lesions observed on MRI during a median follow-up time of 2 years. This inverse longitudinal association between 25(OH)D levels and MS activity was consistent with and independent of HLA-DRB1*15 or vitamin D–binding protein status and was present across various subgroups of patients with different baseline characteristics. A decrease in the number of new active lesions with increasing serum 25(OH)D levels was observed even after excluding patients with vitamin D levels less than 50.0 nmol/L, a result suggesting that 50.0 nmol/L may be too low a target in patients with MS. However, despite the uniform results for MRI measures of disease activity, serum 25(OH)D levels were not significantly correlated with brain atrophy or clinical outcomes.

Strengths of this study include its large number of participants, the regionally diverse population with varying baseline characteristics, and the repeated measurements of 25(OH)D, which helped to characterize patients’ long-term vitamin D status. Furthermore, the results could be adjusted for known confounders, such as baseline disease severity, which affects both serum 25(OH)D levels and MS progression, and the wide distribution of serum 25(OH)D levels allowed us to assess whether optimal levels for preventing MS progression exceed the moderate levels observed in most previous studies.7,9,10 Our observation of the lowest level of MS activity among patients with serum 25(OH)D levels above 100.0 nmol/L is consistent with the results of a previous investigation in the United States20 and suggests that the 25(OH)D levels in most patients with MS who are not receiving supplemental vitamin D may be suboptimal. Although circulating levels of 25(OH)D of 100.0 nmol/L are achievable in most patients with supplementation of cholecalciferol, 1000 to 4000 IU/d, which is considered a tolerable dose, there is no convincing evidence that increasing 25(OH)D levels above the physiologic limit (approximately 150.0 nmol/L) is beneficial, and it could potentially be detrimental. Considering the results of a randomized trial21 in which higher (14 000 IU/d) doses of cholecalciferol were administered to patients with MS, intake of more than 4000 IU/d should be discouraged.

Our study also has important limitations. Length of follow-up was relatively short (median, 2 years). This limited follow-up may explain the lack of association between serum 25(OH)D levels and measures of brain atrophy or clinical end points, both of which have been previously shown to be modified by 25(OH)D in a study with a longer follow-up period.7 However, although the lack of observed effect for measures of brain volume may be the result of a relatively short follow-up time in our study, since other studies including longer follow-up have observed a protective effect of vitamin D, we cannot rule out that brain atrophy may be less sensitive to 25(OH)D levels than the development of new lesions. In addition, participants were at different stages in the disease course, yet the range of disease duration at recruitment in this study was relatively short (median duration, 3 years; interquartile range, 1-7 years). Exposure to UV light is also an important determinant of serum 25(OH)D levels11,22,23 and has immunologic effects not mediated by vitamin D that could potentially affect MS outcomes.22 Nonetheless, the inverse association between serum 25(OH)D levels and MS outcomes in regions with low levels of UV light during most of the year and in patients with more severe disease who are less likely to spend time outdoors support a role for vitamin D. All patients in the present study received interferon beta-1b, and, although uniform treatment is an important advantage, our results may not be generalizable to patients who receive other drugs. Finally, as in all observational studies, confounding by an unmeasured factor related to both serum 25(OH)D levels and MS progression cannot be excluded. However, adjustment for known confounders did not materially influence effect estimates.

Results of this study are consistent with the known immunomodulatory role of vitamin D and also support evidence from clinical trials2429 that have shown dose-dependent effects of vitamin D on the immune system, including a reduced proliferation by peripheral blood T cells in response to MS-related antigens and an increased proportion of interleukin 10+ CD4+ T cells.

Several ongoing randomized trials are assessing the effect of supplementary vitamin D in patients with relapsing-remitting MS and will provide important evidence in establishing causality. Ethical concerns, however, prevent the randomization to placebo of patients with vitamin D insufficiency and, thus, the experimental demonstration of the effects of vitamin D supplementation in these patients. Therefore, our study provides crucial complementary evidence on the importance of correcting vitamin D insufficiency in MS and, in addition, contributes to identifying the optimal 25(OH)D level.

Conclusions

The findings of this large prospective investigation comprising patients from 5 continents and a broad range of serum 25(OH)D levels suggest that adequate vitamin D status is an important determinant of MS activity not only early in the disease course but also several years after the diagnosis. These results also suggest that individuals with MS may benefit from vitamin D levels above those currently considered by many to be sufficient in healthy adults.

Back to top
Article Information
 

†Died August 28, 2015.

Corresponding Author: Kathryn C. Fitzgerald, ScM, Department of Nutrition, Harvard T. H. Chan School of Public Health, Harvard University, 655 Huntington Ave, Bldg 2, Room 347a, Boston, MA 02115 (hpkcf@channing.harvard.edu).

Accepted for Publication: August 10, 2015.

Published Online: October 12, 2015. doi:10.1001/jamaneurol.2015.2742.

Author Contributions: Drs Pohl and Ascherio contributed equally to this work. Ms Fitzgerald and Dr Ascherio had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Pohl, Ascherio.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Fitzgerald.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Fitzgerald, Munger, Köchert, Ascherio.

Obtained funding: Munger, Ascherio.

Administrative, technical, or material support: Köchert, Suarez, Pohl.

Study supervision: Comi, Cook, Hartung, O’Connor, Suarez, Sandbrink, Pohl.

Conflict of Interest Disclosures: Dr Ascherio reported receiving honoraria for speaking at scientific symposia by Almirall, Roche, Sanofi, and Serono. Dr Arnason reported serving as a consultant for Bayer HealthCare Pharmaceuticals. Dr Cook reported receiving honoraria from Bayer HealthCare Pharmaceuticals and from other pharmaceutical companies (Genmab, Merck, and Merck Serono) for speaking and consulting. Dr Comi reported receiving personal compensation for activities with Bayer-Schering, Biogen-Dompé, Merck Serono, Novartis, Sanofi Pharmaceuticals, and Teva Neuroscience and as a consultant. Dr Filippi reported being the editor-in-chief of the Journal of Neurology; serving on scientific advisory boards for Teva Pharmaceutical Industries; receiving compensation for consulting services and/or speaking activities from Biogen Idec, Excemed, Novartis, and Teva Pharmaceutical Industries; and receiving research support from Alzheimer’s Drug Discovery Foundation, Biogen Idec, Cure PSP, Fondazione Italian Ministry of Health, Fondazione Italiana Sclerosi Multipla, Italiana di Ricerca per la SLA, the Jacques and Gloria Gossweiler Foundation (Switzerland), Novartis, and Teva Pharmaceutical Industries. Dr Goodin reported previously or currently participating in several industry-sponsored clinical trials on MS; the sponsoring pharmaceutical companies for these trials include Ares-Serono, Bayer Schering HealthCare, Berlex Laboratories, Biogen Idec, Merck Serono, Novartis, Schering AG, and Teva Neuroscience. In addition, Dr Goodin reported lecturing at medical conferences and in public on various aspects of the epidemiology, diagnosis, and management of MS; in many cases, these talks have been sponsored directly or indirectly by one of the above-listed companies. He reported serving as a temporary ad hoc consultant to several of these organizations. Dr Hartung reported receiving honoraria for consulting and speaking at symposia from Bayer Schering Pharma, Biogen Idec, Genzyme, Merck Serono, Novartis, Roche, Sanofi, and Teva Pharmaceutical Industries, with approval by the rector of Heinrich Heine University. Dr Kappos reported that his institution, the University Hospital Basel, has received research support and payments that were used for research support for Dr Kappos’s activities as principal investigator and member or chair of planning and steering committees or advisory boards in trials sponsored by Actelion, Addex, Bayer HealthCare Pharmaceuticals, Bayer Schering Pharma, Biogen Idec, CLC Behring, Genentech, GeNeuro SA, Genzyme, Merck Serono, Mitsubishi Pharma, Novartis, Octapharma, Praxicon, Roche, Sanofi, Santhera, Siemens, and Teva Pharmaceutical Industries. Dr Kappos reported receiving grants from the European Union, Gianni Rubatto Foundation, the Novartis and Roche Research foundations, the Swiss MS Society, and the Swiss National Research Foundation. Dr Jeffery reported speaking and consulting for Acorda, Alkermes, Bayer, Biogen, Genentec, Genzyme, GlaxoSmithKline, Malinckrodt, Novartis, Pfizer, Serono, and Teva Pharmaceutical Industries. He also reported receiving research support from Bayer, Biogen, Genentec, Genzyme, Novartis, Pfizer, Serono, and Teva Pharmaceutical Industries. Dr O’Connor reported receiving either personal compensation (for consultation, service on a scientific advisory board, or speaking) or financial support for scholarly activities from pharmaceutical companies that develop products for MS, including Abbott Laboratories, Bayer, Biogen Idec, Bio-MS, EMD Serono, Genentech, Genzyme, Novartis, Roche, Sanofi, and Teva Pharmaceutical Industries. Drs Sandbrink and Pohl reported owning stock in Bayer AG, the owner of Bayer Pharma AG and Bayer HealthCare Pharmaceuticals. No other disclosures were reported.

Funding/Support: This study was supported by the National Multiple Sclerosis Society (Dr Ascherio, NMSS RG487515/1). The BEYOND study was sponsored by Bayer.

Role of the Funder/Sponsor: The funding agencies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Previous Presentation: An abstract of this work was given as an oral presentation at the annual meeting of the joint American Committee for Treatment and Research in Multiple Sclerosis–European Committee for Treatment and Research in Multiple Sclerosis; September 13, 2014; Boston, Massachusetts.

References
1.
Rudick  RA, Miller  D, Hass  S,  et al; AFFIRM and SENTINEL Investigators.  Health-related quality of life in multiple sclerosis: effects of natalizumab. Ann Neurol. 2007;62(4):335-346.
PubMedArticle
2.
Frohman  EM, Racke  MK, Raine  CS.  Multiple sclerosis—the plaque and its pathogenesis. N Engl J Med. 2006;354(9):942-955.
PubMedArticle
3.
Munger  KL, Zhang  SM, O’Reilly  E,  et al.  Vitamin D intake and incidence of multiple sclerosis. Neurology. 2004;62(1):60-65.
PubMedArticle
4.
Munger  KL, Levin  LI, Hollis  BW, Howard  NS, Ascherio  A.  Serum 25-hydroxyvitamin D levels and risk of multiple sclerosis. JAMA. 2006;296(23):2832-2838.
PubMedArticle
5.
Banwell  B, Bar-Or  A, Arnold  DL,  et al.  Clinical, environmental, and genetic determinants of multiple sclerosis in children with acute demyelination: a prospective national cohort study. Lancet Neurol. 2011;10(5):436-445.
PubMedArticle
6.
Salzer  J, Hallmans  G, Nyström  M, Stenlund  H, Wadell  G, Sundström  P.  Vitamin D as a protective factor in multiple sclerosis. Neurology. 2012;79(21):2140-2145.
PubMedArticle
7.
Ascherio  A, Munger  KL, White  R,  et al.  Vitamin D as an early predictor of multiple sclerosis activity and progression. JAMA Neurol. 2014;71(3):306-314.
PubMedArticle
8.
Martinelli  V, Dalla Costa  G, Colombo  B,  et al.  Vitamin D levels and risk of multiple sclerosis in patients with clinically isolated syndromes. Mult Scler. 2014;20(2):147-155.
PubMedArticle
9.
Simpson  S  Jr, Taylor  B, Blizzard  L,  et al.  Higher 25-hydroxyvitamin D is associated with lower relapse risk in multiple sclerosis. Ann Neurol. 2010;68(2):193-203.
PubMed
10.
Runia  TF, Hop  WC, de Rijke  YB, Buljevac  D, Hintzen  RQ.  Lower serum vitamin D levels are associated with a higher relapse risk in multiple sclerosis. Neurology. 2012;79(3):261-266.
PubMedArticle
11.
Løken-Amsrud  KI, Holmøy  T, Bakke  SJ,  et al.  Vitamin D and disease activity in multiple sclerosis before and during interferon-β treatment. Neurology. 2012;79(3):267-273.
PubMedArticle
12.
Mowry  EM, Krupp  LB, Milazzo  M,  et al.  Vitamin D status is associated with relapse rate in pediatric-onset multiple sclerosis. Ann Neurol. 2010;67(5):618-624.
PubMed
13.
Devonshire  V, Havrdova  E, Radue  EW,  et al; FREEDOMS Study Group.  Relapse and disability outcomes in patients with multiple sclerosis treated with fingolimod: subgroup analyses of the double-blind, randomised, placebo-controlled FREEDOMS study. Lancet Neurol. 2012;11(5):420-428.
PubMedArticle
14.
O’Connor  P, Filippi  M, Arnason  B,  et al; BEYOND Study Group.  250 µg or 500 µg Interferon beta-1b versus 20 mg glatiramer acetate in relapsing-remitting multiple sclerosis: a prospective, randomised, multicentre study. Lancet Neurol. 2009;8(10):889-897.
PubMedArticle
15.
Lampl  C, Nagl  S, Arnason  B,  et al.  Efficacy and safety of interferon beta-1b sc in older RRMS patients—a post hoc analysis of the BEYOND study. J Neurol. 2013;260(7):1838-1845.
PubMedArticle
16.
Smith  SM, Zhang  Y, Jenkinson  M,  et al.  Accurate, robust, and automated longitudinal and cross-sectional brain change analysis. Neuroimage. 2002;17(1):479-489.
PubMedArticle
17.
Li  Y, Willer  CJ, Ding  J, Scheet  P, Abecasis  GR.  MaCH: using sequence and genotype data to estimate haplotypes and unobserved genotypes. Genet Epidemiol. 2010;34(8):816-834.
PubMedArticle
18.
Wang  YC, Meyerson  L, Tang  YQ, Qian  N.  Statistical methods for the analysis of relapse data in MS clinical trials. J Neurol Sci. 2009;285(1-2):206-211.
PubMedArticle
19.
Andersen  PK, Gill  RD.  Cox’s regression model counting process: a large sample study. Ann Stat. 1982;10:1100-1120.Article
20.
Mowry  EM, Waubant  E, McCulloch  CE,  et al.  Vitamin D status predicts new brain magnetic resonance imaging activity in multiple sclerosis. Ann Neurol. 2012;72(2):234-240.
PubMedArticle
21.
ClinicalTrials.gov. Supplementation of VigantOL Oil Versus Placebo as Add-on in Patients With Relapsing Remitting Multiple Sclerosis Receiving Rebif Treatment (SOLAR). NCT01285401. https://clinicaltrials.gov/ct2/show/NCT01285401. Accessed August 31, 2015.
22.
Becklund  BR, Severson  KS, Vang  SV, DeLuca  HF.  UV radiation suppresses experimental autoimmune encephalomyelitis independent of vitamin D production. Proc Natl Acad Sci U S A. 2010;107(14):6418-6423.
PubMedArticle
23.
Bertrand  KA, Giovannucci  E, Liu  Y,  et al.  Determinants of plasma 25-hydroxyvitamin D and development of prediction models in three US cohorts. Br J Nutr. 2012;108(10):1889-1896.
PubMedArticle
24.
Bock  G, Prietl  B, Mader  JK,  et al.  The effect of vitamin D supplementation on peripheral regulatory T cells and β cell function in healthy humans: a randomized controlled trial. Diabetes Metab Res Rev. 2011;27(8):942-945.
PubMedArticle
25.
Mahon  BD, Gordon  SA, Cruz  J, Cosman  F, Cantorna  MT.  Cytokine profile in patients with multiple sclerosis following vitamin D supplementation. J Neuroimmunol. 2003;134(1-2):128-132.
PubMedArticle
26.
Burton  JM, Kimball  S, Vieth  R,  et al.  A phase I/II dose-escalation trial of vitamin D3 and calcium in multiple sclerosis. Neurology. 2010;74(23):1852-1859.
PubMedArticle
27.
Smolders  J, Peelen  E, Thewissen  M,  et al.  Safety and T cell modulating effects of high dose vitamin D3 supplementation in multiple sclerosis. PLoS One. 2010;5(12):e15235. doi:10.1371/journal.pone.0015235.
PubMedArticle
28.
Jorde  R, Sneve  M, Torjesen  PA, Figenschau  Y, Gøransson  LG, Omdal  R.  No effect of supplementation with cholecalciferol on cytokines and markers of inflammation in overweight and obese subjects. Cytokine. 2010;50(2):175-180.
PubMedArticle
29.
Schleithoff  SS, Zittermann  A, Tenderich  G, Berthold  HK, Stehle  P, Koerfer  R.  Vitamin D supplementation improves cytokine profiles in patients with congestive heart failure: a double-blind, randomized, placebo-controlled trial. Am J Clin Nutr. 2006;83(4):754-759.
PubMed
×