There is disagreement in the literature as to whether e-cigarettes are a gateway to traditional cigarette smoking among adolescents, or whether e-cigarettes displace smoking. Observable data (e.g. from the United States) find a strong displacement effect but many published studies suggest that e-cigarettes act as a gateway to cigarette use by documenting a positive association between e-cigarette use initiation and subsequent cigarette use. Yet, these associations may not be causal and instead due to a confounding factor (i.e. an omitted variable or common liability) that determines both e-cigarette use and subsequent cigarette use. This presentation will try to address this problem of confounding, using a three-pronged approach: smoking rates among adolescents with initial e-cigarette use will be compared with 1) a real-world control group (adolescents with initial non-combustible, non-cigarette tobacco use), 2) a synthetic control group, selected using propensity score matching (PSM) and 3) associations of changes in youth e-cigarette use and smoking uptake at population level will be assessed using timeseries analysis. Data for the first two analyses come from the National Youth Tobacco Survey (2014-2015, N=37,417, age 9-18) and the first three waves of the PATH study (2013-2016, N=5,544, age 12-17) and for the third analysis from the Smoking Toolkit Study (STS) (2008-2018, N=37,105, age 16-24). Results highlight that the impact of e-cigarette use on subsequent smoking is complex and may depend on underlying sample characteristics but that the net effect of e-cigarettes on smoking uptake appears to be negligible at population level.

Presentation slides:
Disentangling the youth gateway of e-cigarettes: findings from three different data sources