A Composite Risk Score for Multiple Sclerosis in First-Degree Relatives: Integrating Genetics and Lifestyle Exposures
Multiple sclerosis (MS) is widely understood to emerge from an interplay between genetic susceptibility and environmental exposures rather than a single precipitating cause. The study by Dobson and colleagues addresses a practical translational challenge: how to identify asymptomatic individuals at meaningfully elevated risk so that pre-symptomatic cohorts can be enriched for longitudinal observation and, ultimately, prevention-oriented trials. This motivation is strengthened by recognition of subclinical stages such as radiologically isolated syndrome (RIS), in which MRI abnormalities consistent with demyelination may precede symptoms and sometimes evolve toward clinical disease. The authors focus specifically on siblings of people with MS, a group known to carry higher risk and to show a higher prevalence of MS-like biomarkers than the general population, making them a rational target population for risk stratification.
Cohort Design and Multimodal Data Collection
The investigators enrolled 302 participants: 78 probands with MS, 121 unaffected siblings, and 103 healthy controls without first- or second-degree relatives with MS, with controls matched to siblings by sex and age decade.
Participants underwent structured clinical history and examination, alongside serological assays and genotyping. Serology included anti–EBNA-1 IgG (as a marker of Epstein–Barr virus immune response), cotinine testing to support smoking exposure classification, and serum 25-hydroxyvitamin D (25-OHvD) quantified by LC–MS/MS and deseasonalised to the day of sampling.
Genetic profiling was performed using the Illumina Immunochip, allowing extraction of MS-associated variants, including HLA-related and non-MHC single nucleotide polymorphisms (SNPs) catalogued from prior genome-wide association studies.
Risk Model Construction: Additive Integration of Genetic and Environmental Effects
A central methodological choice was to translate known risk factors into a single scalar score by summing log-transformed odds ratios in an additive model, following established approaches for risk-score assembly.
Environmental and clinical covariates were weighted using published effect estimates, including female sex (RR 2.22), prior infectious mononucleosis (RR 2.17), smoking exposure (“ever smoking,” RR 1.52), and EBNA-1 IgG quintiles in which the highest titre category carried substantially increased risk (odds ratio 9.4 for Q5 versus reference).
Vitamin D was incorporated using quintiles from prior literature, with only the highest quintile (Q5) providing a statistically meaningful reduction in risk (RR 0.38 versus lowest quintile).
For genetics, two versions were evaluated: (i) a model including HLA-DRB1*1501 only (heterozygote RR 3.1; homozygote RR 6.2), and (ii) a broader model incorporating MS-associated SNPs from the 2011 GWAS rather than the larger 2013 set, to avoid genetic effects numerically dominating environmental contributions under the simple additive framework.
Performance and Group Separation: Evidence of an “Intermediate” Sibling Risk Distribution
Across models, the score separated MS cases from controls and placed unaffected siblings between them, consistent with the hypothesis that siblings constitute an intermediate-risk group. Using the most informative specification—full 2011 GWAS genetic information and excluding serum 25-OHvD due to supplementation-related distortion in the MS group —mean scores were 10.02 in MS, 9.00 in siblings, and 8.20 in healthy controls, with receiver operating characteristic performance reaching an AUC of 0.82 (95% CI 0.75–0.88).
The authors further operationalised the score for trial planning by dividing sibling scores into septiles: the overall difference between siblings and controls was driven largely by the highest sibling septile, implying that approximately 10% of siblings could be prioritised for intensive surveillance in enriched pre-symptomatic studies.
Validation via MRI and Implications for Prevention-Adjacent Research
To test whether high-risk siblings exhibited more subclinical CNS pathology, the study conducted brain MRI (T2/PD imaging) and single-voxel proton spectroscopy in 10 high-score and 10 low-score siblings using blinded assessment and established interpretive criteria.
One high-risk participant developed clinically definite MS before imaging, but among those scanned, lesion number and distribution did not differ materially between high- and low-risk groups; additionally, spectroscopy in normal-appearing white matter showed no detectable between-group differences in metabolites such as N-acetyl-aspartate–related measures or myo-inositol.
The authors interpret these findings cautiously: MRI may be an inefficient primary screening approach at scale, while a composite risk score may be a cost-effective gatekeeper for longitudinal monitoring; however, definitive validation requires prospective follow-up to observe radiological and clinical conversion. They also highlight likely avenues for improving discrimination, including explicit modelling of gene–environment interactions, incorporation of additional exposures (e.g., childhood obesity, which was excluded here due to measurement limitations), and consideration of rare variants not captured by common SNP arrays.
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:
Dobson, R., Ramagopalan, S., Topping, J., Smith, P., Solanky, B., Schmierer, K., ... & Giovannoni, G. (2016). A risk score for predicting multiple sclerosis. PLoS One, 11(11), e0164992.
