Why a 3x Smoking Risk Still Looks Like a Decline
Picture a chart: vaping among teens climbing sharply, while smoking drops. To the casual observer—or even to a policymaker under pressure—it might look like good news. Maybe vaping is helping young people quit cigarettes, or keeping them from starting at all. That’s the “displacement effect” story, and it’s a tempting one.
But according to new research published in Public Health Research & Practice, that interpretation can be dangerously misleading. The study uses a simple but powerful hypothetical model to show how coinciding trends—rising vaping, falling smoking—can appear under very different realities, including ones where vaping actually increases the risk of smoking.
In other words: what you see on the graph may not be the truth that matters.
Two Competing Stories
Public health debates about e-cigarettes often hinge on two opposing ideas:
- The “gateway effect”: Young people who vape are more likely to start smoking later.
- The “displacement effect”: Vaping replaces smoking, reducing cigarette uptake.
At the individual level, evidence for the gateway effect is strong. Meta-analyses of prospective cohort studies consistently show that youth who vape are about three times more likely to start smoking than those who never vape.
But when researchers zoom out to population trends, some see a different picture. Repeated cross-sectional studies from countries like the U.S. and New Zealand have noted that as vaping rose, smoking fell—and concluded that vaping might be pushing smoking out of the picture.
The new study argues this is a classic case of mistaking correlation for causation.
Three Scenarios, Same Graph
To make the point, the authors built a hypothetical example with three steady-state populations of 10,000 adolescents each, starting with 15% smoking prevalence and no vaping in Year 1. Over five years, vaping rises by 3% annually, while smoking among non-vapers declines 10% each year.
They modeled three realities:
- Gateway Effect (RR = 3.0): Vapers are three times as likely to smoke as non-vapers.
- No Effect (RR = 1.0): Vaping doesn’t influence smoking risk at all.
- Displacement Effect (RR = 0.33): Vapers are one-third as likely to smoke.
Surprisingly, in all three cases, the population chart looks similar: vaping goes up, smoking goes down. Even in the gateway scenario—where vaping actively increases smoking risk—the overall smoking rate still declines.
The takeaway? A falling smoking rate doesn’t prove vaping is protective?
Why This Matters for Policy
This isn’t just an academic point. Misinterpretations of such trends have already shaped legislation. The authors note that a 2020 New Zealand study suggesting vaping displaced smoking was cited 61 times in five years, including by British American Tobacco. It became the most frequently referenced evidence in parliamentary submissions on e-cigarette regulation—used to argue against stricter rules.
Without careful statistical analysis, public health policy risks being steered by misleading narratives, sometimes amplified by industries with vested interests.
The Missing “Before and After”
One key problem with the “vaping displaces smoking” studies: they often start tracking after vaping is already common. In New Zealand, for example, by the time data collection began in 2014, about 20% of adolescents had already tried vaping.
A better approach is interrupted time series analysis, which looks at smoking trends before and after vaping’s introduction. That way, researchers can see whether declines in smoking sped up, slowed down, or continued unchanged after vaping became widespread.
Without that “before” picture, it’s impossible to know if vaping caused the decline—or if the drop was already in motion.
A More Careful Reading of the Data
The study’s message is clear: Coinciding trends are not proof of cause and effect. Whether you see vaping and smoking moving in opposite directions or together, it tells you nothing definitive about the relationship between the two.
For researchers, that means building models and analyses that account for pre-existing trends and potential confounders. For policymakers, it means resisting the lure of easy visual narratives and instead demanding stronger evidence before making decisions that affect public health.
What’s Next?
Moving forward, the authors call for:
- More before-and-after studies on the impact of vaping introductions.
- Greater caution in interpreting repeated cross-sectional data.
- Awareness of how misread evidence can be leveraged in policy debates—especially by tobacco industry players.
The vaping–smoking debate is far from settled, but one thing is certain: sloppy interpretations help no one, and could harm the very populations public health is trying to protect.
Join the Conversation
- How should researchers balance the need for timely policy input with the risk of misinterpretation?
- Have you seen other examples where public health policy was swayed by weak or misused data?
- What additional evidence would you want before concluding vaping displaces smoking?
