Vape restrictions may push some young people towards cigarettes, major analysis finds

Vape restrictions may push some young people towards cigarettes, according to a study examining the unintended consequences of policies designed to curb vaping.

Researchers analysing 11 US studies found that measures reducing access to vapes were, overall, associated with increased smoking among young people.

The findings raise questions about judging the success of vape restrictions solely by whether vaping falls, without checking what happens to cigarette use.

“The evidence overall supports the conclusion that e-cigarette availability reduces youth cigarette smoking,” the authors wrote.

The paper is by Christopher Lowenstein and Michael F. Pesko, from the University of Missouri, and Jamie Hartmann-Boyce, from the University of Massachusetts Amherst.

What happened when access changed?

The researchers examined seven studies of minimum legal sales ages for vapes, three of vape taxes and one of flavour restrictions.

Most focused on school-aged teenagers, although the evidence also included young adults and pregnant teenagers.

The studies used policy changes as “natural experiments”, comparing smoking outcomes in populations affected by restrictions with those in comparison groups. This approach can help researchers assess whether changing access to vapes affects smoking.

Combining the results required an extra step. An age restriction, a tax rise and a flavour restriction affect access differently, so the team put them on a common scale based on how much each reduced vaping.

In the main analysis, a restriction equivalent in its effect on vaping to a $1 vape tax increase was associated with a 5.2 per cent relative increase in smoking.

That’s a percentage increase from the underlying smoking rate, rather than a rise of 5.2 percentage points. It’s also a standardised estimate across policies, rather than a prediction that every restriction would have the same effect.

Results strengthened after excluding weaker studies

The researchers checked their findings using a second statistical model that allows more variation between studies.

That model pointed towards a 3.7 per cent increase in smoking, but the uncertainty was wide enough to include a possible decrease. The result was therefore not statistically conclusive.

However, after removing two studies rated at critical risk of bias, both models found a statistically significant 6.8 per cent relative increase in smoking on the same standardised scale.

The main analysis also continued to show a statistically significant increase when the researchers removed each estimate in turn, suggesting that its result did not depend on any single estimate.

“Our meta-analysis of quasi-experimental studies uncovered an inverse association between e-cigarette availability and smoking among young people,” the authors wrote.

In plain terms, greater access to vapes was associated with less smoking, while restricting access was associated with more.

A challenge to the gateway argument

The findings differ from studies that follow individual young people and find that those who vape are more likely to smoke later.

The authors argue that shared influences, including willingness to take risks, friendship groups and family circumstances, may make some young people more likely to try both products. That can make it difficult to establish whether vaping itself causes later smoking.

They also point out that research focused solely on people who have never smoked cannot capture reductions in smoking among existing smokers.

“In contrast, the reduced form studies in our review allow for the potential positive and negative effects of e-cigarette use on smoking at the population-level,” they wrote.

Smoking must remain part of the picture

The findings underline why policies aimed at reducing youth vaping also need to track cigarette use. A reduction in vaping alone cannot establish whether a restriction is achieving its intended public health benefit.

The association with increased smoking became clearer when the researchers excluded the two studies at greatest risk of bias. Both statistical models then produced the same estimate: a 6.8 per cent relative increase in smoking on the study’s standardised policy scale.

The analysis drew on US studies identified through April 2023, all with some risk of bias. Evidence on flavour restrictions was limited to one study, which found no clear change in smoking.