Vitamin D and Multiple Sclerosis: What Genetic Evidence Reveals About Causality
Multiple sclerosis (MS) is a chronic autoimmune and neurodegenerative disorder in which immune-mediated injury to myelin and neural tissue produces sensory, motor, cognitive, and visual impairment. The disease commonly begins in early adulthood, making the identification of modifiable risk factors particularly important. Epidemiological observations have long suggested a connection between vitamin D status and MS: disease prevalence tends to increase at higher latitudes, where ultraviolet radiation and cutaneous vitamin D synthesis are reduced, while lower circulating concentrations of 25-hydroxyvitamin D, or 25OHD, have been associated with increased MS incidence, relapse frequency, magnetic resonance imaging activity, and disability. However, these observations cannot by themselves demonstrate causality. Individuals with early or established MS may spend less time outdoors, resulting in reduced sunlight exposure and lower vitamin D concentrations—a form of reverse causation. Vitamin D status is also correlated with physical activity, adiposity, diet, socioeconomic conditions, and other variables that may confound conventional epidemiological associations. Mokry and colleagues therefore investigated whether genetically determined reductions in 25OHD increase susceptibility to MS, rather than merely accompanying the disease.
Mendelian Randomization as a Tool for Causal Inference
The investigators used Mendelian randomization, an analytical framework in which genetic variants associated with an exposure are employed as instrumental variables to estimate the exposure’s causal effect on an outcome. Because alleles are allocated during meiosis before disease development, genotype is generally less affected by behavioral, environmental, and clinical confounders than a biomarker measured during adulthood. Genetic variants associated with lower vitamin D concentrations can therefore serve as proxies for lifelong differences in vitamin D status. A valid Mendelian randomization instrument must satisfy three central assumptions: it must be reliably associated with the exposure, it must not be systematically associated with confounders, and it must influence the outcome primarily through the exposure rather than through an independent biological pathway. This design is especially relevant to MS because pathological processes may precede clinical diagnosis by many years. A blood measurement obtained after symptom development may consequently reflect disease-related behavioral changes, whereas inherited genetic variation represents an exposure established from conception and maintained throughout life. Mendelian randomization does not reproduce every feature of a randomized clinical trial, but it provides a powerful intermediate form of evidence when long-term prevention trials are unavailable.
Genetic Instruments and Study Populations
Four single-nucleotide polymorphisms were selected from the SUNLIGHT genome-wide association study, which included 33,996 individuals and was, at the time, the largest genetic investigation of circulating 25OHD. The selected variants were rs2282679 in or near GC, rs12785878 near DHCR7, rs10741657 near CYP2R1, and rs6013897 in CYP24A1. These loci are biologically compelling because they represent distinct components of vitamin D physiology. DHCR7 influences the availability of 7-dehydrocholesterol, the cutaneous precursor used in vitamin D synthesis; CYP2R1 encodes a hepatic 25-hydroxylase involved in producing circulating 25OHD; GC encodes the vitamin D-binding protein responsible for transporting most circulating vitamin D metabolites; and CYP24A1 participates in vitamin D catabolism and inactivation. The effects of these variants on 25OHD were estimated in 2,347 participants from the Canadian Multicentre Osteoporosis Study. Their associations with MS were then examined using data from major international genetic consortia containing as many as 14,498 MS cases and 24,091 controls. The pathway diagram on page 8 of the article illustrates how the four loci collectively capture vitamin D synthesis, transport, activation, and metabolism rather than a single isolated biochemical process.
Instrument Validation and Analytical Strategy
The four variants were evaluated carefully before being combined into a Mendelian randomization estimate. They showed no meaningful linkage disequilibrium with one another, indicating that the instruments provided largely independent genetic information. The authors also examined whether the variants were associated with potential confounders or alternative biological pathways, including physical activity, sunlight exposure, smoking, alcohol consumption, body mass index, lipid concentrations, inflammatory biomarkers, and socioeconomic characteristics. The total number of 25OHD-decreasing alleles was strongly associated with lower circulating 25OHD in the Canadian cohort, producing an F-statistic of 49.7 and a highly significant allelic-score association. This is important because weak genetic instruments can generate imprecise or biased causal estimates. For each variant, the investigators calculated a ratio estimate comparing its effect on MS risk with its effect on 25OHD, and then pooled the individual estimates using inverse-variance-weighted fixed-effects and random-effects models. They additionally performed sensitivity analyses excluding variants that might be affected by population stratification or horizontal pleiotropy and compared variants related to vitamin D synthesis with those related to transport and metabolism.
Principal Results: Lower Vitamin D and Higher MS Susceptibility
The central result was a substantial association between genetically lowered vitamin D status and MS susceptibility. In the primary fixed-effects analysis, each one-standard-deviation decrease in natural-log-transformed 25OHD was associated with approximately twice the odds of MS, with an odds ratio of 2.02 and a 95% confidence interval of 1.65–2.46. The corresponding random-effects estimate was similar, at 2.07, although its confidence interval was wider. Moderate heterogeneity was observed across the four variants, but its magnitude was difficult to estimate precisely because only four instruments were available. Importantly, the association persisted after the exclusion of the DHCR7 variant, which raised concerns regarding geographical population structure: the resulting odds ratio remained 1.72. Excluding the GC variant, which could theoretically influence immune function through the vitamin D-binding protein independently of circulating vitamin D, also did not materially weaken the result. When the instruments were separated by biological function, both the synthesis-associated variants and the metabolism-associated variants independently supported increased MS risk, with odds ratios of 2.08 and 1.86, respectively. These convergent estimates strengthen the interpretation that the signal reflects vitamin D biology rather than the idiosyncratic effect of one genetic locus.
Biological and Clinical Interpretation
The findings are biologically plausible because vitamin D is not merely a regulator of calcium homeostasis; its active metabolites also influence innate and adaptive immune responses. Vitamin D receptor signaling can affect antigen presentation, T-cell differentiation, cytokine production, and the transcriptional regulation of immune-related genes, including major histocompatibility complex class II genes implicated in MS susceptibility. Nevertheless, the magnitude of the Mendelian randomization estimate should not be interpreted as a direct prediction of the effect of taking a specific vitamin D supplement. The genetic instruments represent small differences in vitamin D physiology operating continuously across the life course, whereas supplementation usually begins later, lasts for a limited period, and may have different effects according to baseline deficiency, age, genotype, disease stage, and dose. The study translated a one-standard-deviation increase into illustrative 25OHD changes—for example, from 25 to approximately 36.9 nmol/L or from 50 to approximately 73.7 nmol/L—but these calculations are statistical equivalences rather than clinical treatment targets. They do not establish an optimal serum concentration, a safe dose, or a guaranteed reduction in an individual’s risk.
Strengths, Limitations, and Future Research
The study’s principal strengths include the use of large genetic consortia, biologically distinct vitamin D instruments, strong instrument–exposure associations, and multiple sensitivity analyses designed to address confounding, reverse causation, pleiotropy, and population stratification. Its limitations are equally important. Residual horizontal pleiotropy cannot be excluded completely, and the limited number of variants restricts the precision of heterogeneity tests and prevents the application of several modern pleiotropy-robust methods. The participants were predominantly of European ancestry, limiting immediate generalization to populations with different genetic structures, vitamin D metabolism, pigmentation, environmental exposures, or baseline MS incidence. The analysis addresses susceptibility to developing MS and cannot establish that vitamin D modifies relapse rate, disability progression, or treatment response after disease onset. Mendelian randomization also estimates the consequence of lifelong genetically influenced exposure and therefore cannot determine the timing, dose, formulation, or duration of an effective intervention. The article consequently provides strong causal support for low 25OHD as an MS susceptibility factor, but it does not constitute proof that supplementation will prevent or treat the disease. Long-term randomized trials, particularly among vitamin D-deficient individuals and people at elevated MS risk, remain necessary to translate this genetic evidence into clinical and public-health recommendations.
Disclaimer: This blog post is based on the provided research article and is intended for informational purposes only. It is not intended to provide medical advice. Please consult with a healthcare professional for any health concerns.
References:
Mokry, L. E., Ross, S., Ahmad, O. S., Forgetta, V., Smith, G. D., Leong, A., ... & Richards, J. B. (2015). Vitamin D and risk of multiple sclerosis: a Mendelian randomization study. PLoS medicine, 12(8), e1001866.
