Modeling study puts youth vaping lifetime mortality effect between 0 and 31 days

A modeling study estimates that e-cigarette use among Americans who were 12 years old in 2016 changes lifetime premature mortality across that birth cohort by at most 0.44 percentage points, or about 31 days of life. The study appeared in the American Journal of Preventive Medicine; Vaping Post published an account of it.

The underlying document is a paper by Warner, Sun, Brouwer and Méndez in the American Journal of Preventive Medicine. The Vaping Post article describes the methods, results and author-acknowledged limitations of that simulation; it does not present new survey data.

The model follows one birth cohort from age 12

The researchers simulated the lifetime trajectory of everyone in the United States who was 12 years old in 2016, a group the article puts at 4,139,869 people. The cohort was aged 14–15 during the 2018–2019 peak in youth vaping, according to the report.

The simulation relied on National Youth Tobacco Survey data and followed the cohort from age 12 in 2016 to age 17 in 2021. It compared a real-world scenario in which e-cigarettes exist with a counterfactual scenario in which vaping never existed.

Vaping-linked smoker estimate is 25,039 above the no-vaping case

By age 20 in 2024, the real-world scenario produced 141,057 smokers and 441,393 vapers who had never smoked, the article states. Without e-cigarettes, the model estimated 116,018 smokers. The 25,039 difference mixes a possible gateway effect with young people who would have smoked anyway, and the researchers say they cannot separate the two.

Assumed vaping risks drive the range of outcomes

Long-term vaping mortality data do not exist, so the researchers varied four parameters and tested every combination. A parameter is an input value that the model holds fixed within a single run.

Smoking mortality risk came from a long-term US cohort study by age and years since cessation, and the spontaneous smoking cessation rate was set at 5.5% per year. Each scenario was repeated 10 million times per individual to stabilize the estimates.

Least favorable assumptions add 0.44 points; most favorable cut 0.06

With vaping assumed 30% as risky as smoking, no cessation benefit and slow vaping cessation, the probability of premature death rose 0.44 percentage points. With vaping at 5% of smoking risk, a 30% cessation benefit and vaping cessation four times faster, the probability fell 0.06 points.

Expressed in life expectancy, the scenarios produced between 0 and 31 days of life lost, and some assumptions produced a small gain. The Vaping Post account attributes the small net effect to three forces largely canceling: vaping-linked smoking uptake, smoking cessation helped by vaping, and the many vapers who never smoked.

At the individual level, a smoker in the model had roughly a 7–8% risk of premature death. A never-smoking vaper had a risk of 0.15% to 0.84%, depending on assumptions, which the article describes as 10 to 50 times lower.

The authors list limitations that mostly favor vaping

The model ignores non-fatal respiratory and cardiovascular disease and treats every vaper as an intensive user. It uses a broad definition of regular smoker, applies the lowest smoking initiation rate to the whole counterfactual population, and does not reflect device changes between 2016 and 2024. The authors state that most of these biases run against vaping, so a corrected model would probably be no less favorable.

The authors conclude that adolescent vaping in this cohort is very unlikely to be an important source of lifetime premature mortality. Those conclusions depend on assumed vaping risks rather than measured long-term outcomes, and the article reports them as model output.

Retailers, regulators and analysts get a range, not a measured risk

The paper bears on the US tobacco policy debate over youth vaping, because it quantifies how much population-level mortality could shift under a spread of assumptions. Adults who never smoked make up 441,393 of the modeled cohort by 2024, and that group carries the lowest modeled risk. The 25,039 figure and the 0–31 day range are the numbers most likely to be cited in submissions and press coverage.

What the documents leave open

The model stops at mortality, so non-fatal outcomes in this cohort remain unquantified. The model also does not account for nicotine delivery changes in devices sold after 2024, which the Vaping Post report flags as an unaddressed factor.

Questions readers ask

What did the 2016 birth cohort study find about youth vaping and lifetime mortality?

The simulation estimated that e-cigarette use across Americans who were 12 in 2016 changed lifetime premature mortality by between 0.44 percentage points higher and 0.06 points lower, depending on assumptions. In life expectancy terms the effect ranged from 0 to 31 days lost, with some assumptions producing a small gain.

How many smokers did the model attribute to vaping?

The real-world scenario produced 141,057 smokers by age 20 in 2024, against 116,018 in the counterfactual without e-cigarettes, a difference of 25,039. The article states this figure combines a possible gateway effect with young people who would have smoked regardless, which the researchers say they cannot separate.

What vaping mortality risk did the researchers assume?

Because long-term vaping mortality data are lacking, the researchers tested vaping risk at 5%, 15% and 30% of smoking risk, with an extra 50% run as a precaution that experts consulted considered unrealistic. They also varied vaping cessation rates, the added smoking-cessation benefit and years of vaping after quitting cigarettes.

What limitations do the study authors acknowledge?

The authors say the model ignores non-fatal respiratory and cardiovascular disease, treats all vapers as intensive users, uses a broad smoker definition and does not account for device changes between 2016 and 2024. They state most of these biases run against vaping, so a corrected model would probably be no less favorable.