Association of Smoking Behavior, Intensity, and Time Since Cessation with Epigenetic Aging Biomarkers: Results from NHANES 1999-2002

BackgroundSmoking is a major preventable risk factor for all-cause mortality and disability worldwide. It leads to age-related diseases, but the effects and reversibility of smoking behaviors on different epigenetic clocks are not fully explored. ObjectiveTo characterize the association of epigenetic age acceleration in whole blood with active and secondhand smoking (SHS), smoking intensity, and time since cessation among U.S. adults. MethodsThis is a cross-sectional study in adults from the NHANES 1999-2002 survey cycles, a population-based survey representative of the U.S. adult civilian non-institutionalized population. We analyzed 2,320 adults aged [≥]50 years, including non-Hispanic White, non-Hispanic Black, Mexican American, and other populations. Those without available self-reported smoking status data were excluded. Smoking exposure was analyzed in terms of self-reported smoking status (current, former, never), intensity (packs in the last month), years since smoking cessation, and SHS (serum cotinine levels: 0.05-10 ng/ml). Epigenetic age was estimated using 12 DNA methylation age biomarkers. Survey-weighted linear models were used to estimate the association of smoking exposure with epigenetic age acceleration while adjusting for confounders and multiple comparisons. ResultsWe analyzed 1,043 never, 903 former, and 374 current smokers (mean age: 65.1{+/-}9.3 years, female: 49.1%). GrimAge2 was 9.1 years (95% CI: 8.0, 10.2) and 2.8 years (95% CI: 2.3, 3.3) higher in current and former smokers, respectively, than in never smokers. Smokers showed an increased pace of aging, with current smokers aging 0.15 (95% CI: 0.13, 0.17) and former smokers 0.04 (95% CI: 0.03, 0.05) additional years per chronological year, and shorter methylation-predicted telomere length (current: -132{+/-}19 bp; former: -30{+/-}15 bp). Each cigarette pack smoked in the past month was associated with increases of 0.1 years in GrimAge2 and PhenoAge, and 0.01 aged months/year in aging pace. Among former smokers, each year since smoking cessation was associated with a deceleration of -0.14 (GrimAge2) and -0.06 (PhenoAge) years, and -0.03 aged months/year in aging pace. Cotinine analyses supported dose-dependent associations of epigenetic aging with smoking and suggested a 0.8-year increase in GrimAge2 with SHS exposure. ConclusionsSmoking was associated in a dose-dependent manner with accelerated epigenetic aging in former and current smokers. However, epigenetic age acceleration declines with time since smoking cessation among former smokers.