Predicting risk of lung function impairment and all-cause mortality using a DNA methylation-based classifier of tobacco smoke exposure

Christina M Eckhardt; Haotian Wu; Diddier Prada; Pantel S Vokonas; David Sparrow; Lifang Hou; Joel Schwartz; Andrea A Baccarelli

doi:10.1016/j.rmed.2022.106896

Predicting risk of lung function impairment and all-cause mortality using a DNA methylation-based classifier of tobacco smoke exposure

Respir Med. 2022 Aug-Sep:200:106896. doi: 10.1016/j.rmed.2022.106896. Epub 2022 Jun 2.

Authors

Christina M Eckhardt¹, Haotian Wu², Diddier Prada³, Pantel S Vokonas⁴, David Sparrow⁴, Lifang Hou⁵, Joel Schwartz⁶, Andrea A Baccarelli²

Affiliations

¹ Columbia University Irving Medical Center, Division of Pulmonary, Allergy and Critical, Care Medicine, Department of Medicine, New York, NY, USA. Electronic address: Cme2113@cumc.columbia.edu.
² Columbia University Mailman School of Public Health, Environmental Health Sciences, Department, New York, NY, USA.
³ Columbia University Mailman School of Public Health, Environmental Health Sciences, Department, New York, NY, USA; Instituto Nacional de Cancerología, México City, Mexico.
⁴ Boston University School of Medicine, VA Normative Aging Study, VA, Boston, USA; Healthcare System and Department of Medicine, Boston, MA, USA.
⁵ Northwestern Feinberg School of Medicine, Department of Preventive Medicine, Chicago, IL, USA.
⁶ Harvard T.H. Chan School of Public Health, Department of Epidemiology, Cambridge, MA, USA.

Abstract

Background: The Epigenetic Smoking Status Estimator (EpiSmokEr) predicts smoking phenotypes based on DNA methylation at 121 CpG sites.

Objective: Evaluate associations of EpiSmokEr-predicted versus self-reported smoking phenotypes with lung function and all-cause mortality in a cohort of older adults.

Methods: The prospective Normative Aging Study collected DNA methylation measurements from 1999 to 2012 with follow-up through 2016. The R package EpiSmokEr derived predicted smoking phenotypes based on DNA methylation levels assayed by the Illumina HumanMethylation450 Beadchip. Spirometry was collected every 3-5 years. Airflow limitation was defined as forced expiratory volume in 1 s/forced vital capacity <0.7. Vital status was monitored through periodic mailings.

Results: Among 784 participants contributing 5414 person-years of follow-up, the EpiSmokEr-predicted smoking phenotypes matched the self-reported phenotypes for 228 (97%) never smokers and 22 (71%) current smokers. In contrast, EpiSmokEr classified 407 (79%) self-reported former smokers as never smokers. Nonetheless, the EpiSmokEr-predicted former smoking phenotype was more strongly associated with incident airflow limitation (hazard ratio [HR] = 3.15, 95% confidence interval [CI] = 1.50-6.59) and mortality (HR = 2.11, 95% CI = 1.56-2.85) compared to the self-reported former smoking phenotype (airflow limitation: HR = 2.21, 95% CI = 1.13-4.33; mortality: HR = 1.08, 95% CI = 0.86-1.36). Risk of airflow limitation and death did not differ among self-reported never smokers and former smokers who were classified as never smokers. The discriminative accuracy of EpiSmokEr-predicted phenotypes for incident airflow limitation and mortality was improved compared to self-reported phenotypes.

Conclusions: The DNA methylation-based EpiSmokEr classifier may be a useful surrogate of smoking-induced lung damage and may identify former smokers most at risk of adverse smoking-related health effects.

Keywords: All-cause mortality; DNA Methylation; Lung function; Smoking.

Publication types

Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov't

MeSH terms

DNA Methylation / genetics
Forced Expiratory Volume
Humans
Lung
Prospective Studies
Risk Factors
Tobacco Smoke Pollution*

Substances

Tobacco Smoke Pollution

Abstract

Publication types

MeSH terms

Substances

Grants and funding