In 3,562 participants from the U.S. Health and Retirement Study (mean age 69.6 years), childhood smoking was associated with 0.65 years of accelerated GrimAge—a mortality-predictive epigenetic clock—while each additional year of education was linked to a 0.10-year deceleration. Across 731,474 methylation sites, education mapped to 574 distinct CpG sites (including cg07318158 in OTUD7B, β = −0.47), childhood smoking to 20 sites enriched for nervous system development pathways, rural upbringing to 49 sites, and grandparent cohabitation to 23 sites. Rural living and grandparent cohabitation showed no significant GrimAge acceleration.
This large observational study reinforces an emerging paradigm: the epigenome acts as a molecular ledger of early social and environmental exposures, with entries readable 50–60 years later. The GrimAge effect sizes here are modest but biologically plausible—childhood smoking coincides with critical neurodevelopmental windows, consistent with the nervous-system pathway enrichment found. Education's broad methylation footprint across embryonic development genes suggests schooling may influence epigenetic programming beyond cognitive pathways. Crucially, however, this is a preprint not yet peer-reviewed, and causal interpretation remains limited: self-reported childhood exposures introduce recall bias, blood-based methylation may not reflect tissue-specific changes, and residual confounding from adult behaviors is difficult to fully exclude. If replicated, these methylation signatures could serve as early biomarkers to stratify lifetime disease risk—an incremental but clinically meaningful contribution to the life-course epigenetics field.
