AI-powered research aims to protect young adults from tobacco addiction
Every year, hundreds of thousands of kids and young adults take their first puff of an e-cigarette – oblivious that this small choice could lead to a lifetime of tobacco addiction. University of Florida researchers are working to change that by using artificial intelligence to pinpoint which young people are most at risk – before they ever start.
The PreventiGators, a team of graduate students and their faculty mentors from the University of Florida’s Department of Family, Youth and Community Sciences, recently earned national recognition for their use of AI to address tobacco use risk. Their analysis of 3,950 youth and emerging adults used machine learning via UF’s supercomputer, HiPerGator, to predict which U.S. youth might be most vulnerable to becoming tobacco users in adulthood by looking at individual behaviors, community influences and digital environments – including social media use – to predict future behavior.
The team found that teens who spend more time on their phones — especially on social media — and who are exposed to tobacco ads or believe tobacco relieves stress are more likely to use tobacco later in life.
Tobacco use is the leading cause of preventable death in the United States, including deaths due to secondhand smoke, according to the U.S. Centers for Disease Control and Prevention. In 2022, 49.2 million – or nearly 1 in 5 – adults said they used tobacco products.
With the rise of vaping, young adults are more at risk for early exposure to tobacco. According to the CDC, 1 in 10 young adults ages 18 to 20 used electronic cigarettes, and about 1 in 6 adults ages 21 to 24 used them, based on a 2023 analysis.
But not all teens are equally vulnerable. Kids who have strong anti-tobacco beliefs or who don’t live with anyone who smokes are less at risk.
“Our goal is to stop tobacco use before it starts,” said Olanrewaju Lawal, a Ph.D. student and team member. “By identifying risk early — especially among youth with no prior tobacco use — we can design smarter, more targeted prevention strategies that protect young people and reflect the digital realities they are navigating today during a critical stage of their development when lifelong health behaviors are being shaped.”
This level of precision would be nearly impossible using traditional statistical methods, said Xiaoya Zhang, family, youth and community sciences assistant professor.
“AI allows us to capture the complexity of real life,” Zhang said. “It helps us understand not just who is at risk, but how — and what we can do about it before it’s too late.”
These prevention-focused researchers hope their analysis – which they hope to publish in a peer-reviewed journal soon – will help create better strategies to protect youth from the long-term effects of tobacco use.
“Tobacco use among youth always ebbs and flows,” said Tennisha Riley, family, youth and community sciences assistant professor. “But prevention science is about staying ahead of those trends. If we can prevent youth from starting in the first place, we don’t have to deal with the devastating consequences later.”
The PreventiGators’ project was selected as the winner of the annual Society for Prevention Research’s Sloboda and Bukoski SPR Cup, a mark of excellence in the prevention science research space.
“This win reflects the growing strength of our academic programs and our commitment to using science and technology to solve real problems,” Riley said. “We’re proud to be training the next generation of prevention scientists here at UF.”