ALS Patients Don't Age Faster Biologically — Except Those With C9orf72 Mutation

ALS is a rapidly fatal neurodegenerative disease and aging is its single strongest environmental risk factor. A leading hypothesis holds that ALS patients undergo accelerated biological aging, measurable through DNA methylation-based 'epigenetic clocks.' Prior studies supported this view, reporting elevated pace-of-aging scores in ALS blood. This study set out to rigorously test that hypothesis using the largest ALS methylation dataset ever assembled, while carefully dissecting potential confounders — most notably, the well-documented shifts in white blood cell (WBC) composition that occur in ALS.

The team analyzed whole-blood methylome profiles from 7,302 individuals — 5,146 ALS patients and 2,156 healthy controls — genotyped on either the Illumina HumanMethylation450k (N=4,425) or HumanMethylationEPIC (N=2,877) arrays as part of Project MinE. Biological age was estimated using five established epigenetic clocks spanning three generations: the first-generation Horvath clock, the second-generation mortality-predictive PCGrimAge, and the third-generation pan-mammalian clock. Pace of aging was measured with DunedinPACE and leukocyte telomere length with DNAmTL. Epigenetic age acceleration (EAA) for each individual was calculated as the residual from a Loess regression of DNA methylation age on chronological age, anchored to controls as the reference baseline. Logistic regression models were compared with and without WBC proportions as covariates using AIC model selection and Wilcoxon signed-rank testing.

The central finding is a methodological one with major biological implications: when white blood cell type proportions were included as covariates, the apparent increase in pace of aging in ALS patients versus controls was eliminated across all epigenetic clocks tested. Model fit improved significantly when WBC data were added (lower AIC, p<0.05 by Wilcoxon signed-rank test), confirming that immune cell compositional changes — not biological aging per se — were driving the apparent EAA signal. This is consequential because ALS is known to involve neutrophilia and lymphopenia, which independently alter methylation patterns in ways that mimic aging signals in epigenetic clock algorithms.

The one robust exception emerged in the C9orf72 subgroup analysis. Among ALS patients alone, those carrying the C9orf72 hexanucleotide repeat expansion showed significantly elevated EAA compared to C9orf72-negative ALS patients, even after adjusting for WBC proportions, sex, smoking status, batch, and site of onset. This suggests that the C9orf72 mutation imparts a distinct epigenetic aging burden beyond immune dysregulation, possibly through its known effects on nucleocytoplasmic transport, autophagy, and repeat-associated non-ATG translation. The direction and magnitude of this effect were consistent across clock types.

Despite these case-control differences in the C9orf72 group, survival analysis in 5,022 ALS patients with known survival status found that none of the EAA scores — from any clock — predicted survival time. This null finding is notable and suggests that whatever biological aging signal is present in ALS does not translate into differential disease progression within the patient population. Sensitivity analyses excluding individuals with imputed chronological age (N=6,048) and replacing cell-type covariates yielded consistent results, strengthening confidence in the primary conclusions. The study reframes ALS epigenomics: immune cell dysregulation is the dominant and actionable methylation signal in blood, while true epigenetic age acceleration is specific to the C9orf72 genetic subtype.