Revisiting the Diffusion of Vaping among College Students
The relationship between young adults and nicotine-containing products has evolved considerably over the past half-century. While traditional tobacco cigarettes were once commonly seen right outside the classroom, scientific findings and educational practices advance, public health campaigns take hold, stigmas alter, culture changes, and views of the behavior shift. With the downfall of cigarettes, a trend toward e-cigarettes followed with widespread commercial production and popularity since the early 2000s [1]. More specifically, with the rise of well-marketed, sleeker, lightweight, and ostensibly less harmful cigarette alternatives, a considerable rise in pod vaping (e.g., JUUL) has occurred among young adults – a 250% increase from 6% of college students in 2017 to 15% in 2018 [2]. From a networks’ perspective, the spread of pod vaping can be modeled as a cascade or diffusion of behavior on college campuses; however, with more detailed studies and an increase in public efforts to reduce nicotine addiction in the last several years, there is more to add to the collective commentary on the diffusion of vaping mentioned on occasion in this blog. Vaping behavior is indeed more dynamic than one may think. Of significance and insight on diffusion networks, a recent focus group study and analysis this year offers an in-depth investigation of vaping use innovations among college students with a KAP (knowledge, attitudes, practices) approach drawn from diffusion of innovations theory [3].
In an obvious sense, college students are likely subject to above average levels of peer pressure due to their moderate density ‘friend group’ structure (p ≈ 2/3, possibly), and high spatial clustering on campus (dorm-living), creating more (albeit weak) ties. Additionally, at a pivotal point in personal development, these individuals may be subject to high adoption rates of new social phenomena—perhaps a ‘friend group’ cluster contains a threshold value of q ≈ 1/3. Evidently, though these particular values are contrived, even without an internal initial adopter, the entire group could be subject to a complete cascade of pod vaping use if p<=1-q for this idealized model, assuming all ‘bridges’ or non-cluster friends adopt the incoming behavior of vaping. Even though friend group clusters are obstacles to cascades, they are not invincible. Particularly if clusters are not universally strong rather than weak ties, they can be overcome by a behavioral cascade. Nonetheless, due to the prevalence of JUUL advertisements in various social media [4], sporadic internal initial adopters are still possible, and in some other settings, almost as likely as ‘irl,’ person-to-person diffusion.
The KAP approach to the analysis contributed key details and additional nuance that I believe can be utilized to better represent the network cascade. Regarding knowledge (K), prominent themes realized in the study included an awareness to addictiveness and uncertainty regarding the health risks. With regard to attitudes (A), a dual perspective was identified, where social use is acceptable, but everyday use is stigmatized. And most consequently, of interest to a hypothetical enhanced cascade model methodology, practices (P) divided well into three evident stages: social connection, wherein pod vaping devices were used to achieve social advantage with peers; addiction, characterized by solitary, habitual use; and detachment, sometimes marked by a dramatic, public action filmed for social media, a mounting trend over roughly the last year and a half [5] [6].
I see such stages of vaping behavior analogous to the three steps of the SIR epidemic model – where any given node is either susceptible, infectious, or removed. Being targeted by peers and social media with conflicting signals and a growing, overwhelming anti-nicotine effort pushed by public health campaigns and organizations recently (e.g., Truth Initiative [7]), the network of young adults and their current diffusion of pod vaping should be modeled by adopting a more “life-cycle”-oriented network structure in addition to its edge-dependent network cascade basis. But would an SIR or SIS (susceptible, infectious, susceptible) epidemiological model more accurately portray reality? Relatedly, how common is a pod vaping/nicotine relapse? For these questions, I am still unsure. I would like to see more data collection and social behavior studies that can improve an approximation for q and expand on the relative influence of young adult social edges in the real vs virtual settings—perhaps one of these ties is stronger than the other.
Links:
- https://casaa.org/education/vaping/historical-timeline-of-electronic-cigarettes/
- Schulenberg JE, Johnston LD, O’Malley PM, Bachman JG, Miech RA, Patrick ME, Accessed 2019. Monitoring the Future national survey results on drug use, 1975–2018: Volume II, College students and adults ages 19–60. Ann Arbor: Institute for Social Research, The University of Michigan. http://monitoringthefuture.org/pubs.html#monographs.
- Katz, S. J., Cohen, E. L., & Kinzer, H. T. (2020). “Can I hit that?” Vaping knowledge, attitudes and practices of college students. Journal of American College Health, 1-10. https://pubmed.ncbi.nlm.nih.gov/33048639/
- https://www.cnn.com/2018/12/17/health/juul-social-media-influencers/index.html
- https://www.theverge.com/2019/9/11/20859645/juul-quit-vaping-cdc-fda-flavor-pod-ban-death
- https://www.youtube.com/watch?v=4nKfunltLxs&ab_channel=truthorange
- https://truthinitiative.org/who-we-are/our-mission