0
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Prospective predictors of electronic nicotine delivery system initiation in tobacco naive young adults: A machine learning approach

      research-article

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          The use of electronic nicotine delivery systems (ENDS) is increasing among young adults. However, there are few studies regarding predictors of ENDS initiation in tobacco-naive young adults. Identifying the risk and protective factors of ENDS initiation that are specific to tobacco-naive young adults will enable the creation of targeted policies and prevention programs. This study used machine learning (ML) to create predictive models, identify risk and protective factors for ENDS initiation for tobacco-naive young adults, and the relationship between these predictors and the prediction of ENDS initiation. We used nationally representative data of tobacco-naive young adults in the U.S drawn from the Population Assessment of Tobacco and Health (PATH) longitudinal cohort survey. Respondents were young adults (18–24 years) who had never used any tobacco products in Wave 4 and who completed Waves 4 and 5 interviews. ML techniques were used to create models and determine predictors at 1-year follow-up from Wave 4 data. Among the 2,746 tobacco-naive young adults at baseline, 309 initiated ENDS use at 1-year follow-up. The top five prospective predictors of ENDS initiation were susceptibility to ENDS, increased days of physical exercise specifically designed to strengthen muscles, frequency of social media use, marijuana use and susceptibility to cigarettes. This study identified previously unreported and emerging predictors of ENDS initiation that warrant further investigation and provided comprehensive information on the predictors of ENDS initiation. Furthermore, this study showed that ML is a promising technique that can aid ENDS monitoring and prevention programs.

          Related collections

          Most cited references41

          • Record: found
          • Abstract: found
          • Article: found
          Is Open Access

          The Precision-Recall Plot Is More Informative than the ROC Plot When Evaluating Binary Classifiers on Imbalanced Datasets

          Binary classifiers are routinely evaluated with performance measures such as sensitivity and specificity, and performance is frequently illustrated with Receiver Operating Characteristics (ROC) plots. Alternative measures such as positive predictive value (PPV) and the associated Precision/Recall (PRC) plots are used less frequently. Many bioinformatics studies develop and evaluate classifiers that are to be applied to strongly imbalanced datasets in which the number of negatives outweighs the number of positives significantly. While ROC plots are visually appealing and provide an overview of a classifier's performance across a wide range of specificities, one can ask whether ROC plots could be misleading when applied in imbalanced classification scenarios. We show here that the visual interpretability of ROC plots in the context of imbalanced datasets can be deceptive with respect to conclusions about the reliability of classification performance, owing to an intuitive but wrong interpretation of specificity. PRC plots, on the other hand, can provide the viewer with an accurate prediction of future classification performance due to the fact that they evaluate the fraction of true positives among positive predictions. Our findings have potential implications for the interpretation of a large number of studies that use ROC plots on imbalanced datasets.
            Bookmark
            • Record: found
            • Abstract: not found
            • Article: not found

            Points of Significance: Statistics versus machine learning

              Bookmark
              • Record: found
              • Abstract: found
              • Article: found
              Is Open Access

              Tobacco Product Use Among Adults — United States, 2019

              Cigarette smoking remains the leading cause of preventable disease and death in the United States ( 1 ). The prevalence of current cigarette smoking among U.S. adults has declined over the past several decades, with a prevalence of 13.7% in 2018 ( 2 ). However, a variety of combustible, noncombustible, and electronic tobacco products are available in the United States ( 1 , 3 ). To assess recent national estimates of tobacco product use among U.S. adults aged ≥18 years, CDC analyzed data from the 2019 National Health Interview Survey (NHIS). In 2019, an estimated 50.6 million U.S. adults (20.8%) reported currently using any tobacco product, including cigarettes (14.0%), e-cigarettes (4.5%), cigars (3.6%), smokeless tobacco (2.4%), and pipes* (1.0%). † Most current tobacco product users (80.5%) reported using combustible products (cigarettes, cigars, or pipes), and 18.6% reported using two or more tobacco products. § The prevalence of any current tobacco product use was higher among males; adults aged ≤65 years; non-Hispanic American Indian/Alaska Native (AI/AN) adults; those whose highest level of educational attainment was a General Educational Development (GED) certificate; those with an annual household income 30% or unweighted denominator 30% or unweighted denominator <50. The figure is a bar chart showing the cigarette smoking status (current, former, or never) among current adult e-cigarette users, by age group. The prevalence of any current tobacco product use was higher among males (26.2%) than among females (15.7%) and among those aged 25–44 years (25.3%), 45–64 years (23.0%), or 18–24 years (18.2%) than among those aged ≥65 years (11.4%) (Table). Current tobacco product use was also higher among non-Hispanic AI/AN adults (29.3%), non-Hispanic adults of other †††† races (28.1%), non-Hispanic White adults (23.3%), non-Hispanic Black adults (20.7%), and Hispanic or Latino adults (13.2%) than among non-Hispanic Asian adults (11.0%); and among those living in the Midwest (23.7%) or South (22.9%) than among those in the Northeast (18.5%) or West (16.4%). The prevalence of current tobacco product use was higher among those whose highest educational attainment was a GED (43.7%) than among those with other levels of education; among those who were divorced/separated/widowed (23.5%) or single/never married/not living with a partner (23.0%) than among those married/living with a partner (19.2%); among those who had annual household income of <$35,000 (27.0%) than among those with higher income; and among LGB adults (29.9%) than among those who were heterosexual/straight (20.5%). Prevalence was also higher among adults who were uninsured (30.2%), insured by Medicaid (30.0%), or had some other public insurance (25.6%) than among those with private insurance (18.0%) or Medicare only (11.4%); among those who had a disability (26.9%) compared with those without (20.1%); and among those who had GAD-7 scores indicating mild (30.4%), moderate (34.2%) or severe (45.3%) anxiety than among those indicating no or minimal (18.4%) anxiety. Discussion In 2019, approximately one in five U.S. adults (50.6 million) reported currently using any tobacco product. Cigarettes were the most commonly used tobacco product among adults, and combustible tobacco products (cigarettes, cigars, or pipes) were used by most (80.5%) adult tobacco product users. Most of the death and disease from tobacco use in the United States is primarily caused by cigarettes and other combustible products ( 1 ); therefore, continued efforts to reduce all forms of combustible tobacco smoking among U.S. adults are warranted. Moreover, approximately one in five current tobacco product users (18.6%) reported using two or more tobacco products, and differences in prevalence of tobacco use were also seen across population groups, with higher prevalence among those with a GED, American Indian/Alaska Natives, uninsured adults and adults with Medicaid, and LGB adults. Each of these groups has experienced social, economic, and environmental stressors that might contribute to higher tobacco use prevalence ( 6 ). Comprehensive strategies at the national, state, and local levels, including targeted interventions and tailored community engagement, can reduce tobacco-related disease and death and help to mitigate tobacco-related disparities ( 1 , 4 , 6 ). U.S. adults also reported using various noncigarette tobacco products, with e-cigarettes being the most commonly used noncigarette tobacco product (4.5%). E-cigarette use was highest among adults aged 18–24 years (9.3%), with over half (56.0%) of these young adults reporting that they had never smoked cigarettes. In addition, the tobacco product with the highest percentage of users aged 18–24 years (24.5%) was e-cigarettes. E-cigarettes contain nicotine, which is highly addictive, can prime the brain for addiction to other drugs, and can harm brain development, which continues until about age 25 years ( 3 ). Although e-cigarette use was lower among the older age groups, more than 40% of e-cigarette users in the 25–44, 45–64 and ≥65 years age groups reported being former smokers. Although some evidence suggests that the use of e-cigarettes containing nicotine and more frequent use of e-cigarettes are associated with increased smoking cessation, smokers need to completely stop smoking cigarettes and stop using any other tobacco product to achieve meaningful health benefits ( 6 , 7 ). The U.S. Surgeon General concluded that there is presently inadequate evidence to conclude that e-cigarettes, in general, increase smoking cessation, and further research is needed on the effects that e-cigarettes have on cessation ( 7 ). Therefore, continued efforts to reduce use of all tobacco products, combustible and noncombustible, are needed. The findings in this report are subject to at least four limitations. First, the 59.1% response rate might have resulted in nonresponse bias, although sample weighting is designed to account for this. Second, self-reported responses were not validated by biochemical testing for cotinine (a biomarker indicating nicotine exposure); however, there is high correlation between self-reported smoking and smokeless use and cotinine levels ( 8 , 9 ). Third, because NHIS is limited to the noninstitutionalized U.S. civilian population, these results might not be generalizable to institutionalized populations and persons in the military. Finally, this analysis does not provide comparisons of prevalence estimates with previous surveys because changes in weighting and design methodology for the 2019 NHIS have the potential to affect comparisons of weighted survey estimates over time. §§§§ The implementation of comprehensive, evidence-based, population-level interventions in coordination with regulation of tobacco products, can reduce tobacco-related disease, disparities, and death in the United States ( 1 , 4 ). These evidence-based, population-level strategies include implementation of tobacco price increases, comprehensive smoke-free policies, high-impact antitobacco media campaigns, and barrier-free cessation coverage ( 1 ). As part of a comprehensive approach, targeted interventions are also warranted to reach subpopulations with the highest prevalence of use, which might vary by tobacco product type. Summary What is already known about this topic? Cigarette smoking remains the leading cause of preventable disease and death in the United States; however, a variety of new combustible, noncombustible, and electronic tobacco products are available in the United States. What is added by this report? In 2019, approximately 20.8% of U.S. adults (50.6 million) currently used any tobacco product. Cigarettes were the most commonly used tobacco product among adults, and e-cigarettes were the most commonly used noncigarette tobacco product (4.5%). The highest prevalence of e-cigarette use was among smokers aged 18–24 years (9.3%), with over half (56.0%) of these young adults reporting that they had never smoked cigarettes. What are the implications for public health practice? The implementation of comprehensive, evidence-based, population-level interventions, combined with targeted strategies, in coordination with regulation of tobacco products, can reduce tobacco-related disease and death in the United States. As part of a comprehensive approach, targeted interventions are also warranted to reach subpopulations with the greatest use, which might vary by tobacco product type.
                Bookmark

                Author and article information

                Contributors
                Journal
                Prev Med Rep
                Preventive Medicine Reports
                2211-3355
                13 February 2023
                April 2023
                13 February 2023
                : 32
                : 102148
                Affiliations
                [a ]Department of Medicine, University of Connecticut School of Medicine, Farmington, CT 06030, USA
                [b ]Department of Psychiatry, Yale University School of Medicine, Connecticut Mental Health Center, 34 Park Street, New Haven, CT 06519, USA
                [c ]Laboratory for Systems Medicine, Department of Medicine, University of Florida, Gainesville, FL 32610, USA
                [d ]Division of Behavioral Sciences and Community Health, University of Connecticut Health Center, Farmington, CT 06030, USA
                Author notes
                [* ]Corresponding author at: University of Connecticut, Department of Medicine, 263 Farmington Avenue, Farmington, CT 06030, USA. atuegwu@ 123456uchc.edu
                Article
                S2211-3355(23)00039-6 102148
                10.1016/j.pmedr.2023.102148
                9971268
                36865398
                7bd83d33-7bf3-48f2-9001-861bee73721d
                © 2023 The Authors

                This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

                History
                : 9 October 2022
                : 11 January 2023
                : 10 February 2023
                Categories
                Regular Article

                e-cigarette,electronic nicotine delivery systems,ends,machine learning,tobacco naïve,never tobacco users,vaping,young adults,prospective predictors,path,population assessment of tobacco and health survey

                Comments

                Comment on this article