This study leveraged the All of Us Research Program to examine both phenotypic and genetic differences across five modes of lifetime nicotine use. Existing genetic research has predominantly focused on cigarette smoking, yet patterns of use, user characteristics, and possible environmental exposures differ by mode. The authors aimed to identify shared and mode-specific genetic risk factors for lifetime use of electronic nicotine (e-nicotine) products, cigars, hookah (tobacco in a hookah), smokeless tobacco, and having smoked at least 100 cigarettes in a lifetime.
Analyses were performed in a cross-ancestry sample of N = 304,213 participants from the All of Us Research Program. The five phenotypes were defined as lifetime use of: electronic nicotine products, cigars, hookah tobacco, smokeless tobacco, and lifetime cigarette use defined by at least 100 cigarettes. The study used common-variant genome-wide association study (GWAS) methods and combined results across ancestral groups in meta-analyses.
Across modes, lifetime nicotine users were more likely to be male and younger, except that cigarette users did not follow the younger-age trend. All modes of use were positively associated with diagnoses of tobacco use disorder. These phenotypic patterns underline that user demographics and clinical comorbidity differ by product type, which could reflect distinct environmental exposures and behavioral drivers of use.
Meta-analyses of GWAS summary statistics identified differing numbers of genome-wide significant loci across modes. Specifically, the authors report 87 genome-wide loci for lifetime cigarette use, 4 loci for lifetime e-nicotine product use, 1 locus for cigar use, 1 locus for smokeless tobacco use, and no genome-wide loci for hookah use. The analyses recovered previously reported loci for some phenotypes; for example, CHRNA4 was identified in association with cigarette use. The paper reports these locus counts as evidence of contrast in genetic architecture between modes.
Pairwise genetic correlations revealed two clusters of genetically related modes. Cigarette, e-nicotine, and smokeless tobacco use had strong positive genetic correlations with each other (reported rg between 0.82 and 0.92). Separately, cigar and hookah use were strongly genetically correlated with each other (reported rg = 0.87). These patterns indicate that some modes share substantial genetic liability, while others cluster together, suggesting mode-specific genetic influences.
Cross-trait genetic correlation analyses showed divergent relationships by mode. Cigarette, e-nicotine, and smokeless tobacco use exhibited the strongest positive genetic correlations with psychiatric disorders, as well as with traits such as negative urgency and neuroticism. By contrast, cigar and hookah use showed null-to-weak genetic correlations with psychiatric disorders but positive correlations with sensation-seeking and openness. These contrasts imply that the genetic liability underlying different modes of nicotine use aligns with different behavioral and psychiatric phenotypes.
The collective results highlight meaningful phenotypic and genetic distinctions across modes of nicotine use. Modes commonly grouped together in epidemiologic studies may differ in underlying genetic architecture and in their genetic relationships with psychiatric and personality traits. The authors interpret these findings as evidence that both personality traits (for example, impulsivity-related measures) and socioeconomic context may contribute differently to who uses which nicotine product and to the genetic liability associated with that use. The identification of multiple loci for cigarette use compared with fewer loci for other modes suggests differences in power, phenotype heterogeneity, or true differences in genetic architecture.
The study used secondary data from the All of Us Research Program; the All of Us Research Program obtained participant consent and institutional approvals at the time of data collection. The authors state the analysis of these de-identified secondary data did not require additional IRB approval. All data are available to registered users on the All of Us Researcher Workbench. The authors declared no competing interests.
In summary, cross-ancestry GWAS in a large All of Us sample (N = 304,213) revealed that lifetime nicotine use modes differ in demographic and clinical profiles, genome-wide locus discovery, and genetic correlations with psychiatric and personality traits. Cigarette, e-nicotine, and smokeless tobacco use form one genetically correlated cluster with links to psychiatric traits and negative urgency, while cigar and hookah use form a separate cluster more related to sensation-seeking and openness. These distinctions point to mode-specific genetic and phenotypic etiologies that merit further study.