Polygenic scores for tobacco use provide insights into systemic health risks in a diverse EHR-linked biobank in Los Angeles

Importance Tobacco use is heavily influenced by environmental factors with significant genetic contributions. An extensive evaluation of the genetic variants predisposing to tobacco use is necessary to understand associated health risks and formulate equitable health policies. Objective To evaluate the predictive performance, risk stratification, and potential systemic health effects of tobacco use disorder (TUD) predisposing germline variants using a European-derived polygenic score (PGS) in the UCLA ATLAS biobank. Design, Setting, and Participants Publicly available TUD-PGS, developed in European ancestry individuals, was evaluated in participants enrolled in the UCLA ATLAS biobank - a multi-ancestry, hospital-based biobank with participant genotypes linked to their de-identified medical records. Main Outcomes and Measures The outcomes of interest were (a) tobacco use disorder and (b) 1,847 phenotypes, identified by phecodes (aggregated ICD codes) extracted from electronic health records. Results Among genetically inferred ancestry groups (GIAs), TUD-PGS associated with TUD in European American (EA) (OR: 1.20, CI: [1.16, 1.24]), Hispanic/Latin American (HL) (OR:1.19, CI: [1.11, 1.28]), and East Asian American (EAA) (OR: 1.18, CI: [1.06, 1.31]) GIAs but not in African American (AA) GIA (OR: 1.04, CI: [0.93, 1.17]). Similarly, TUD-PGS offered strong risk stratification across quantiles in the EA and HL GIAs but inconsistently in EAA and AA GIAs. In a cross-ancestry phenome-wide association meta-analysis, TUD-PGS was associated with cardiac, respiratory, psychiatric, and metabolic phecodes (17 phecodes at P < 2.57e-5). In individuals with no history of smoking (never-smokers), the top TUD-PGS associations were obesity and alcohol-related disorders (P = 3.54E-07, 1.61E-06). Mendelian Randomization (MR) analysis provides evidence of a causal association between tobacco use and adiposity measures. Conclusions and Relevance This study provides an investigation of TUD-PGS across multiple ancestries and a range of phenotypes in a hospital-based biobank, suggesting shared biological pathways between tobacco use, alcohol use disorder, and obesity. European ancestry TUD-PGS demonstrated inconsistent performance in non-European ancestries for risk prediction and stratification. Equitable clinical translation of TUD-PGS will require the inclusion of multiple ancestry populations at all levels of TUD genetic research. Additionally, TUD-predisposed individuals may require comprehensive tobacco use management approaches to address underlying addictive tendencies. Key Points Question How does a European-derived polygenic score (PGS) for tobacco use disorder (TUD) perform in ancestrally diverse individuals within a real-world medical system and what are the systemic health risks in TUD predisposed individuals? Findings European ancestry TUD-PGS performs inconsistently in non-European ancestry populations. TUD-PGS correlates with cardiometabolic, respiratory, and psychiatric phenotypes. In individuals with no history of smoking, TUD-PGS correlates with obesity and alcohol-related disorders. Meaning Existing TUD-PGS is not generalizable across ancestries. TUD-predisposed individuals are at risk for obesity and alcohol use disorder in the absence of smoking behavior, suggesting shared genetic etiology between these phenotypes.