A Polygenic Score for Smoking Persistence Captures Differences in Cessation Age

Smoking cessation substantially reduces tobacco-related morbidity and mortality, with earlier cessation conferring greater health benefits. Polygenic scores (PGSs) for smoking persistence may provide insight into cessation timing across diverse populations. This study examined the association between a multi-ancestry PGS for smoking persistence and smoking cessation timing in a genetically diverse population.

This study included 139,129 current and former smokers with genetic data from the All of Us Research Program. A multi-ancestry PGS for smoking persistence was constructed using genome-wide association study summary statistics from the GWAS & Sequencing Consortium of Alcohol and Nicotine Use Phase 2 (GSCAN2). Smoking status and cessation timing were examined across the PGS distribution, with a focus on differences between the highest and lowest deciles. Median cessation ages and cumulative cessation probabilities were estimated, while Cox proportional hazards models compared cessation hazards across PGS groups overall and by genetic ancestry.

Among 139,129 participants, the proportion of current smokers increased across the PGS distribution, from 35.5% in the lowest decile to 51.9% in the highest decile. Unadjusted median cessation ages were 38 and 50 years in the lowest and highest deciles, respectively (N = 125,047). After adjusting for sex, educational attainment, household income, cigarettes smoked per day, and genetic population structure, estimated median cessation ages were 40 years (95% CI: 40-40) and 48 years (95% CI: 47-50), respectively, representing an 8-year difference. Participants in the highest decile also had a lower cessation hazard than those in the lowest decile (HR = 0.67, 95% CI: 0.65-0.70), with associations observed across genetic ancestry groups, although estimates varied in precision.

A multi-ancestry PGS for smoking persistence was associated with substantial differences in cessation timing across genetically diverse populations. These findings suggest that genetic liability to smoking persistence may help characterize risk of delayed cessation and inform earlier, more targeted smoking cessation support.