Duration of smoking cessation and DNA methylation aging clocks among US adults: potential reversibility of smoking-induced biological aging.

Abstract

Tobacco smoking is a critical modifiable risk factor for age-related diseases and mortality. Although smoking cessation is known to improve health outcomes, whether smoking-induced biological aging is reversible upon cessation remains unclear. We investigated the associations of smoking status and cessation duration with epigenetic age acceleration (EAA) using multi-generational DNA methylation clocks. We analyzed data from 1,911 adults aged 50-84 years in the National Health and Nutrition Examination Survey (NHANES) 1999-2002. Participants were classified as never, former, or current smokers, with former smokers further categorized by cessation duration (< 10, 10- < 20, and ≥ 20 years). EAA was evaluated across six validated DNA methylation clocks: first-generation (HorvathAge, HannumAge, SkinBloodAge), second-generation (PhenoAge, GrimAge2), and third-generation (DunedinPoAm). In survey-weighted regression, current smokers exhibited the highest EAA across all clocks. Among former smokers, a longer duration of cessation was associated with a dose-response reduction in EAA, with the strongest associations observed for GrimAge2 and DunedinPoAm. Former smokers with ≥ 20 years of cessation showed a substantial reduction in GrimAge2 acceleration (β = -12.66 years; 95% CI: -14.56, -10.75), approaching the levels observed among never smokers (β = -13.88 years; 95% CI: -15.70, -12.05). Restricted cubic spline analysis revealed rapid initial declines in EAA during the first decade post-cessation, with continuous declines observed beyond 20 years for GrimAge2 and DunedinPoAm. These findings suggest that longer durations of smoking cessation are associated with more favorable biological aging profiles, adding biological evidence for the long-term health benefits of sustained cessation for healthy aging.