Metabolic Clues to Multiple Sclerosis: What Mendelian Randomization Reveals
Multiple sclerosis (MS) is a chronic autoimmune disorder of the central nervous system characterized by neuroinflammation, demyelination, and progressive neurodegeneration. Although genetic susceptibility contributes substantially to disease risk, environmental and metabolic factors—including smoking, Epstein–Barr virus infection, obesity, and vitamin D deficiency—also influence MS development. Metabolomic studies have identified numerous biochemical differences between patients with MS and healthy individuals, particularly in amino-acid, lipid, nucleotide, and energy metabolism. However, conventional case-control studies cannot easily determine whether these metabolic alterations contribute to disease initiation or arise as consequences of inflammation, treatment, dietary change, or neurological deterioration. The study by Ge and colleagues addresses this problem through a metabolome-wide Mendelian randomization analysis designed to prioritize circulating metabolites that may have causal effects on MS susceptibility.
Mendelian Randomization as a Genetic Framework for Causality
Mendelian randomization uses genetic variants associated with an exposure—in this case, circulating metabolite concentrations—as instrumental variables for estimating the exposure’s effect on a disease outcome. Because alleles are allocated at conception, genetic instruments are generally less affected by the socioeconomic, behavioral, clinical, and environmental confounders that complicate observational epidemiology. The approach can also reduce reverse-causation bias because the onset of MS cannot alter an individual’s inherited genotype. Conceptually, this natural allocation resembles certain features of random assignment in a clinical trial, although Mendelian randomization remains dependent on important assumptions: the genetic variants must predict the exposure, must not be associated with relevant confounders, and must influence MS only through the metabolite being investigated. The authors therefore used several complementary analytical methods to test the robustness of their causal estimates and to identify possible violations caused by horizontal pleiotropy.
A Large-Scale Metabolome-Wide Study Design
The investigators integrated summary statistics from three genome-wide association studies of circulating metabolites, involving 7,824, 24,925, and 115,078 participants, respectively. Genetic associations with MS were obtained from an International Multiple Sclerosis Genetics Consortium dataset containing 14,802 cases and 26,703 controls, with all contributing studies conducted in populations of European ancestry. Two thresholds were used to select metabolite-associated single-nucleotide polymorphisms: the conventional genome-wide significance threshold of P< 5×10−8 and a less stringent threshold of P < 1×10−6, which increased metabolome coverage while also requiring more cautious interpretation. After excluding metabolites represented by fewer than three independent genetic instruments, the researchers examined 404 metabolites under the stricter threshold and 571 under the broader threshold. As illustrated by the workflow diagram on page 3, the primary inverse-variance-weighted analysis was supplemented by MR-Egger, weighted-median, weighted-mode, MR-PRESSO, heterogeneity, and MR Steiger analyses.
Lipoprotein Composition Appears More Informative Than Total Lipid Levels
Among the 29 metabolites prioritized as potentially causal, ten represented lipids carried within specific lipoprotein subclasses. The results demonstrate that the biological interpretation of circulating lipids depends not only on lipid identity but also on the size and type of the transporting particle. Genetically predicted increases in total cholesterol, phospholipids, and triglycerides within large very-low-density lipoprotein particles were associated with lower MS risk, with odds ratios of 0.83, 0.80, and 0.81, respectively. Conversely, total cholesterol, phospholipids, and cholesterol esters within very large high-density lipoprotein particles were associated with higher risk. The forest plot on page 4 displays this directional contrast, suggesting that aggregate measurements such as total HDL cholesterol may conceal functionally distinct lipoprotein subclasses. These findings reinforce the importance of high-resolution lipid profiling rather than treating HDL, LDL, or triglycerides as metabolically uniform exposures.
Amino Acids and Ketone Bodies Emerge as Candidate Risk Factors
Several amino acids and energy-related metabolites were also associated with MS susceptibility. A genetically predicted one-standard-deviation increase in circulating serine was associated with a 56% increase in MS odds, whereas lysine and O-sulfo-L-tyrosine were associated with more modest increases. Serine is biologically notable because it contributes to one-carbon metabolism and serves as a precursor for phosphatidylserine and sphingomyelin, lipids that are integral to myelin structure. The authors further note that Epstein–Barr virus can stimulate serine import and biosynthesis in B cells, providing a plausible connection between viral exposure, immune-cell metabolism, and MS risk. Acetoacetate and acetone, two ketone bodies, showed particularly large effect estimates, with odds ratios of approximately 2.47 and 2.45. The scatter, forest, and leave-one-out analyses presented on page 5 indicate that the serine result was not driven by a single genetic variant, while sensitivity analyses supported the ketone-body associations despite greater statistical uncertainty.
Scientific Strengths and Necessary Caution
The study’s principal strength is its systematic evaluation of hundreds of metabolites within a causal-inference framework that is less vulnerable than conventional observational research to residual confounding and reverse causation. The use of two instrument-selection thresholds, strong instruments with F-statistics above 10, and six complementary Mendelian randomization procedures further increases confidence in the overall prioritization strategy. Nevertheless, the findings should not be interpreted as definitive proof that directly modifying these metabolites will prevent MS. Horizontal pleiotropy cannot be completely excluded, some metabolites lacked sufficient genetic instruments, and differences among metabolomic GWAS datasets may have affected statistical power. The analyses also assumed approximately linear exposure–outcome relationships, reflected lifelong genetically influenced metabolite differences rather than short-term interventions, and could not identify critical developmental periods or distinguish among MS subtypes. Furthermore, restriction to individuals of European ancestry limits the immediate generalizability of the results to populations with different genetic backgrounds.
Implications for Biomarker Discovery and Therapeutic Research
This metabolome-wide analysis provides a genetically informed shortlist of biochemical pathways that warrant experimental and clinical investigation. Serine, lysine, acetone, acetoacetate, uridine, gamma-glutamyl amino acids, and subclass-specific lipids may help reveal how systemic metabolism interacts with immune activation, B-cell biology, myelin composition, and neurodegeneration before the clinical onset of MS. However, a causal association with disease risk does not automatically imply that dietary supplementation, pharmacological inhibition, or direct metabolite reduction will produce therapeutic benefit; intervention effects may differ from lifelong genetic effects and may depend on disease stage, tissue compartment, or metabolic context. Future work should therefore combine replication in ancestrally diverse cohorts with multivariable Mendelian randomization, longitudinal metabolomics, cellular studies, animal models, and carefully designed clinical investigations. The study ultimately establishes a valuable framework for converting large metabolomic datasets into biologically prioritized hypotheses for MS prevention, early detection, and therapeutic development.
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:
Ge, A., Sun, Y., Kiker, T., Zhou, Y., & Ye, K. (2023). A metabolome-wide Mendelian randomization study prioritizes potential causal circulating metabolites for multiple sclerosis. Journal of neuroimmunology, 379, 578105.
