2022 Funding Cycle Awardees

View other award years: 2024 • 2022 • 2021 • 2020 • 2019

Please note: Listing describes appointments and affiliations at the time of award. Please check our Trainees and Alumni pages for current appointments and affiliations.

Title: “The Effects of Tobacco 21 Adoption on Cigar Sales in the US”

PI: Dr. Alex Liber, Assistant Professor at Georgetown University

Abstract and accomplishments +

Dr. Alex LiberGeorgetown University

Abstract

Introduction: Laws raising the minimum age of sale to 21 years of age for tobacco products (Tobacco 21 or T21 laws) proliferated across the US in the past decade, leading to the adoption of a federal T21 law in December 2019. By severing the social sources of tobacco products in US high schools, T21 appears to have helped further contain youth tobacco use. Prior research has confirmed that T21 policies have decreased both the use of tobacco products among young people and the sales of the cigarettes which young people smoke. However, little research has yet examined the effects of T21 passage on cigar use, the second-most used tobacco product among young people in the US. Glover-Kudon et al found that sales declines of large cigars and cigarillos declined faster than national trends after Hawaii and California passed the first state-level T21 laws in 2016. Little else has examined whether cigar smoking among young people has been affected by the proliferation of this important health policy. Aims: The study utilized state-level cigar sales data from Nielsen Company to examine trends before and after the implementation of T21 laws and the impact of key provisions such as purchase, use, and possession (PUP) rules. The research also identifies cigar brands that are disproportionately consumed by young people and explores their sales patterns. Additionally, the project includes two spinoff studies: one examining the effects of Modified Risk Tobacco Product (MRTP) marketing authorization on General Snus sales and another assessing the comprehensiveness of tobacco sales data by comparing it to excise tax collections. Results: Preliminary findings indicate that the adoption of T21 laws is associated with a significant decline in sales of cigars predominantly consumed by young people, reinforcing the effectiveness of these policies in reducing youth access to tobacco products. The research also reveals that the MRTP marketing authorization for General Snus led to a modest increase in sales, while other snus brands experienced larger increases. Furthermore, the sales data analysis demonstrates such data's utility as a surveillance and policy evaluation tool for understanding tobacco market dynamics and monitoring the effectiveness of interventions. Implications: This work suggests that T21 policies effectively reduce youth cigar smoking and highlight the need for comprehensive and timely sales data for tobacco control and policy evaluation. The findings also inform the ongoing evaluation of MRTP marketing authorizations and their effects on tobacco product sales. Overall, this project contributes to evidence-based policymaking, advances understanding of the public health implications of the tobacco industry, and enhances research rigor in the field of tobacco regulatory science.

Accomplishments

Fall 2022: This pilot project yielded a National Institutes of Health Loan Repayment Program Award of the same name.

March 2023: Findings from this project were presented at the 2023 Society for Research on Nicotine and Tobacco Annual Meeting.

 

Title: “Predicting Smoking Behaviors Using Machine Learning”

PI: Mona Issabakhsh, Research Instructor at Georgetown University and Thuy Le, Postdoctoral Fellow at University of Michigan

Abstract and accomplishments +

Thuy LeUniversity of Michigan 

Mona IssabakhshGeorgetown University

Abstract

Introduction: The United States’ smoking prevalence has significantly decreased over time (from 23.3% in 2000 to 13.7% in 2018). Cigarette smoking, however, is currently responsible for about 480,000 deaths annually and is still a major public health issue. Identification of factors and policies driving the transition of individuals between never smokers, current smokers, and former smokers is a critical need. Machine learning has been investigated widely in the last decades in various research studies and can recognize patterns and detect complicated relationships among data features, which humans are not able to do, to make accurate estimations, predictions, and decisions. Several studies have recently started to apply machine learning algorithms in tobacco research, such as smoker status classification from narrative clinical texts, and tobacco-related outcome prediction using administrative, survey, or clinical trial data. Current literature on smoking prevalence mainly employs mathematical and statistical models, accounting for few predictors (e.g., age, sex, and race). Complex multistate Markov models are established for smoking prevalence prediction, considering more predictors (e.g., age, sex, race, education, and income). These models focus only on a limited number of factors and policies to explain the transition of individuals between never, current, and former smokers. Another disadvantage of these models is that the state transition rates must be estimated, which increases the complexity of the model development. Machine learning algorithms make use of a flexible model structure with little or no parameter estimations, allowing for rapid updates and modifications. Machine learning also enables more efficient use of massive data for tobacco research by accounting for multiple predictors and policies and the discovery of complex patterns in large datasets to produce high-quality estimations and predictions. Given the enormous data on tobacco use, both cross-sectional and longitudinal, machine learning is a promising tool to leverage all the available data. Methods: For this study, we will apply binary machine learning classifiers to group individuals by smoking status (cigarette users and non-users). Classification is a supervised machine learning methodology that determines which class the dependent variable (response) belongs to, based on one or more independent variables (predictors). Classification algorithms involve predicting a qualitative response for an observation, or in other words, assigning the observation to a category or class. A classification algorithm learns from labeled data. After understanding the data, it determines how to best map input data to specific class labels by associating patterns to the unlabeled new data and learns how to assign labels to the new data. The dataset, therefore, must sufficiently represent the problem and have multiple examples of each class label. Many possible classification techniques or classifiers are available in the literature. Aims: The specific aims of this project are: Aim 1: Develop and train machine learning classifiers using the data from two pairs of PATH waves (1-2 and 4-5) to predict the transition from never smokers to current smokers and identify the key risk factors of smoking initiation. Aim 2: Develop, train, test, and validate machine learning classifiers using the data from two pairs of PATH waves (1-2 and 2-3) to predict the transition from current smokers to former smokers and identify the key factors of smoking cessation. Key Findings: For the transition from never smokers to current smokers: Using PATH waves 1 & 2 and 4 & 5, we discovered important features. Across the considered waves, two factors, (i) BMI and (ii) dental/oral health status, robustly appeared as important predictors of smoking initiation, besides other well-established predictors. For the transition from current to former smokers: Our analysis indicated that a higher degree of past 30 days e-cigarette use, a lower degree of past 30 days cigarette use, ages older than 18 at smoking initiation, fewer years of smoking, poly tobacco past 30-days use (compared with only cigarettes use), and higher BMI mainly resulted in higher chances of cigarette cessation for adult smokers in the US.

Accomplishments

March 2023: Findings from this project were presented at the 2023 Society for Research on Nicotine and Tobacco Annual Meeting.

April 2023: Dr. Le was the lead author and Dr. Issabakhsh the second author of an article resulting from their pilot project work. The article, published in Nicotine & Tobacco Research, is entitled “Are the relevant risk factors being adequately captured in empirical studies of smoking initiation? A machine learning analysis based on the Population Assessment of Tobacco and Health study.”

June 2023: Dr. Issabakhsh was the lead author and Dr. Le a co-author of an article resulting from their pilot project work. The article, published in PLOS One, is entitled “Machine learning application for predicting smoking cessation among US adults: An analysis of waves 1-3 of the PATH study.”