Lipid Genetics and Disability Progression in Multiple Sclerosis: Evidence for Risk-Allele–Dependent Effects of HDL and Cholesterol Balance
Multiple sclerosis (MS) is characterised by substantial heterogeneity in both pathology and clinical course, and while genetic and environmental contributors to disease onset are increasingly well mapped, determinants of disability progression remain comparatively under-resolved. Building on prior observations that serum lipid measures (for example, total cholesterol, high-density lipoprotein (HDL), and the total cholesterol:HDL ratio) and, in some studies, body mass index (BMI) associate with disability levels and trajectories, Zhang and colleagues asked a more mechanistic question: do common lipid-related (and BMI-related) genetic polymorphisms modulate the relationship between these metabolic traits and subsequent disability progression?
Cohort and Outcomes: Prospective Measurement of EDSS Change
The investigators analysed 184 participants from the Ausimmune Longitudinal (AusLong) Study who had definite MS, genome-wide genotyping, and Expanded Disability Status Scale (EDSS) assessments at baseline and at 5-year review. Disability progression was operationalised as the annualised EDSS change (ΔEDSS) over approximately five years, with careful handling of baseline EDSS assumptions for relapse-onset versus progressive-onset disease. The cohort was predominantly female (81%), with a mean age of approximately 38 years, and exhibited a mean annualised disability change of +0.29 EDSS points; baseline lipid and BMI distributions were also reported to contextualise metabolic status at study entry.
Genetic Architecture Considered: From GWAS Hits to Candidate SNP Sets
To compile an a priori set of variants plausibly relevant to lipid traits and BMI, the authors performed a structured literature review of genome-wide association studies (GWAS), ultimately selecting three lipid GWAS and one BMI GWAS. From these sources, they assembled 162 lipid-associated single-nucleotide polymorphisms (SNPs) and 97 BMI-associated SNPs, using directly genotyped variants where available and proxy variants otherwise. Genotyping used the Illumina Human Exome BeadChip platform (with an MS-relevant customised component), and associations with ΔEDSS were tested using linear regression adjusted for key covariates (age, sex, study site, and relapse status near the 5-year EDSS assessment), alongside multiple-testing control and permutation analyses to evaluate robustness.
Principal Genetic Findings: Five Lipid SNPs and a Risk-Allele Dose Effect
Among the lipid SNP set, five variants were nominally associated with ΔEDSS (p< 0.05): rs2013208 (nearest gene RBM5), rs9488822 (FRK), rs17173637 (TMEM176A), rs10401969 (CILP2), and rs2277862 (ERGIC3). Although none survived stringent multiple-testing correction individually, the authors observed suggestive allele dose-dependency for several loci and proceeded to aggregate these signals using a cumulative genetic risk score (CGRS). The lipid CGRS demonstrated a strong dose–response association with disability progression (ptrend=1.4×10⁻⁶): participants with ≥6 risk alleles progressed approximately 0.38 EDSS points per year faster than those with ≤3 risk alleles, and the CGRS model explained 16% of the variance in ΔEDSS after covariate adjustment.
Gene–Environment Interaction: Lipid Levels Matter Most in Genetically Susceptible Individuals
A central contribution of the paper is the explicit testing of interaction between baseline lipid measures and the lipid CGRS in predicting ΔEDSS. Significant interactions were reported for HDL (pinteraction=0.005) and for the total cholesterol:HDL ratio (pinteraction=0.030): as illustrated in the figure (page 4), individuals with more than four risk alleles showed markedly higher predicted disability progression when HDL was lower or when the cholesterol:HDL ratio was higher, whereas those with ≤4 risk alleles exhibited comparatively little separation across lipid strata. Importantly, the combined models (CGRS plus lipid parameter) explained substantially more disability variance than lipid measures alone—26% for HDL and 27% for the cholesterol:HDL ratio—consistent with a meaningful gene–environment synergy rather than simple additive effects.
BMI Genetics: Minimal Signal Relative to Lipid Genetics
In contrast to the lipid findings, BMI-related genetic variation showed limited evidence of relevance to disability change in this dataset. Only one BMI-associated SNP, rs2033529 (nearest gene LRFN2), was nominally associated with ΔEDSS, and there was no significant interaction between this variant and baseline BMI in predicting disability progression. The study also reported no significant change in BMI over the 5-year follow-up and, notably, no detectable relationship between changes in lipid variables over time and ΔEDSS, nor interactions between lipid changes and the CGRS—suggesting that baseline status and genetic susceptibility were more informative in this analysis than longitudinal metabolic drift.
Interpretation and Implications: Toward Risk-Stratified Metabolic Intervention in MS
The authors interpret these results as evidence that common lipid-related variants, especially in aggregate, may influence early MS disability accrual and may amplify the clinical relevance of an adverse lipid profile. They also acknowledge complexity: two of the five lipid SNPs showed directions of effect on disability that were not straightforwardly aligned with their lipid-trait directions in GWAS, raising possibilities including residual type I error, non-linear epistasis within the CGRS, or pleiotropic biology beyond lipid metabolism. Strengths include the prospective design, clinically meaningful follow-up duration, and attempts to mitigate false discovery via permutation and CGRS-based inference; the principal limitation is sample size, which constrains power after multiple-testing correction and underscores the need for replication in independent longitudinal MS cohorts. Clinically, the findings offer a plausible framework for explaining heterogeneous outcomes in statin trials in MS—namely, that lipid-lowering or lipid-modifying interventions may preferentially benefit individuals with a susceptible genetic profile—thereby motivating future work on biomarker-driven stratification and mechanistic validation of the implicated loci and pathways.
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:
Zhang, Y., Zhou, Y., Van Der Mei, I. A., Simpson, S., Ponsonby, A. L., Lucas, R. M., ... & Taylor, B. V. (2019). Lipid-related genetic polymorphisms significantly modulate the association between lipids and disability progression in multiple sclerosis. Journal of Neurology, Neurosurgery & Psychiatry, 90(6), 636-641.
