The Population Is Shifting to a Less Harmful Form of Nicotine

My new paper shows that US adults are in the midst of switching from combustible cigarettes to e-cigarette use, with little dual use. Many other countries show similar transitions.

Artwork by Pawel Bugucki

A new paper with colleagues Sooyong Kim (Pinney Associates) and Prof. Ray Niaura (NYU) presents a straightforward depiction of prevalence trends for cigarette smoking and e-cigarette use in US adults.

Before I get into the results, I want to note that our article, published in Nicotine & Tobacco Research, was a personal side project (not commissioned, funded, or overseen by my clients Juul Labs or the Global Forum on Nicotine). In this paper we analyze the National Health Interview Survey, which is a nationally-representative annual survey conducted by the CDC. Data are also publicly available, so anyone who doesn’t trust our analysis is welcome to redo our analysis.

While there are several official publications and fact sheets from the CDC presenting prevalence trends, we improved on previous publications in 3 ways:

1. Combined smoking-vaping status, rather than looking at each product in isolation (which doesn’t distinguish between exclusive vs dual use) — which Brad Rodu also recently did (Rodu & Plurphanswat 2024).

2. Finer-grained categories: separately tracking some day vs. every day use, rather than treating all past-month use in the same category — which Sarah Jackson and colleagues did recently with England’s Smoking Toolkit Study (Jackson et al., 2024).

3. Using quarterly rather than annual data, which makes results more amenable to formal trend analysis and modeling — also similar to the Smoking Toolkit Study.

Shift in Product Use, but Not in Overall Nicotine Demand

First we looked at trends in the percentage of US adults (18+) who used either cigarettes or e-cigarettes. (This doesn’t include other products such as cigars, smokeless, or pouches, but since these are a lot less common, we don’t expect this simplification to affect the results.)

What’s remarkable is how stable this is over time, especially considering the drastic changes under the hood. Over 2014-2018, the use of either product remained approximately stable at 16-17% of the US adult population. I wonder if this is the natural “floor” once all the conventional tobacco control measures have been exhausted (educating people about the harms, age restrictions, marketing restrictions, taxes, outdoor smoking bans, etc.).

But within this overall stability of either smoking or vaping, consumption shifted dramatically:

• Substantial declines in cigarette smoking (red = daily smoking; orange = nondaily smoking)

• Substantial increases in e-cigarette use (shades of green for daily and nondaily vaping)

• Dual use (shades of blue) had a slight U-shaped curve, but was fairly uncommon throughout this period (~10-15% of people who use either product).

• Even within dual use, there is a modest shift over time from more harmful subtypes of dual use (daily smoking with either daily or nondaily vaping) to less harmful subtypes (involving nondaily smoking).

I should note that this is normalized within those who either smoke or vape, so each bar here goes inside the corresponding pink bar above. But since the pink bars above are approximately stable throughout the time period, the bars below are approximately comparable over time. (Whereas if the proportion of nicotine users (pink bars above), say, doubled or halved over the time period, it would complicate the comparisons.)

Legend: Dual Use Behaviors: Dual-DD = Dual use of daily smoking and daily vaping; Dual-DND = Dual use of daily smoking and non-daily vaping; Daul-NDD = Dual use of non-daily smoking and daily vaping; Dual-NDND = Dual use of non-daily smoking and non-daily vaping; Exclusive Smoking/Vaping: D = Daily; ND = Non-daily

Next, we looked at how these findings varied by age. There was strong patterning by age, where the changes were much more prominent in younger age groups, whereas older adults had much more stable use over time (which unfortunately means stagnant cigarette smoking rates).

A “Reality Check” on Concerns about Dual Use

The main conclusion from our analysis is that, over the past 10 years, the US population has shifted from the most harmful form of nicotine consumption to a much lower-risk one, all while the overall demand for nicotine has remained approximately stable.

The other prominent conclusion is about dual use, which is the current battleground for tobacco harm reduction research. Critics such as Stanton Glantz claim that dual use poses higher risks than even exclusive smoking and criticize e-cigarettes on the basis that:

providing e-cigarettes as a cessation aid produced 2.3 dual users for every person who stopped smoking cigarettes. Glantz, 2026

If you’re not familiar with the dual-use debate, please see my recent post going into the scientific flaws in general and Glantz’s piece specifically.

The claim that using e-cigarettes creates more dual users than “switchers” (and presumably the critics would extend this not just to medical use of e-cigarettes as a cessation aid but perhaps even especially to consumer use as an alternative nicotine product) would mean that a large majority of e-cigarette users dual-use with cigarettes rather than exclusively vape.

But our new paper gives an important “reality check” to that. The majority of e-cigarette users exclusively vape. Not only is dual use actually quite uncommon (~10-15% in the plots above), but dual use is not becoming much more common over time. A little, perhaps, given the U-shaped trend, but it’s modest and involves a shift to the less-harmful subtypes of dual use.

I use the phrase “reality check” as an Easter egg referring to Prof. David Levy’s landmark 2019 paper offering strong evidence opposing the e-cigarette gateway hypothesis. The parallels are strong: both involve an alarmist theory that predicts e-cigarettes causing a large increase in cigarette smoking, but in both cases this is contradicted by actual population-level data.

Here, there are nowhere near as many dual users predicted by the “2.3 dual users for every switcher” claim. Not only is dual use the minority behavior, but it is not even growing / scaling with e-cigarette use: dual use only changed by ~50% at most (from ~10 to 15%) while e-cigarette use grew several fold.

A Remarkably Similar Pattern across Countries

Comparisons across countries are often tricky, but are interesting to look at with respect to similar dynamics. The reason I’m bringing this up is because it offers a counterargument to people who would look at my paper above and say “well, cigarette smoking was already declining, and the uptake of vaping is opposing that success.” I believe strongly that is a misinterpretation; I believe smoking would not have declined so drastically if there weren’t a different nicotine product to use instead.

This is a question of a counterfactual (what would have happened if e-cigarettes hadn’t come along?) which is hard to study. But looking at other countries with different prevalence, regulation, and product categories can show that similar dynamics occur when a population shifts from one category to another. This is not an exhaustive list, but I will add to it if readers point out examples I’m missing.

Norway: Snus displaced smoked tobacco starting in the 1990s, first for males and then for females. The transition is largely complete now without any increase in total nicotine use (Lund & Lund, 2014).

Sweden: Like Norway, the population largely shifted away from smoked tobacco to snus (Norberg et al. 2011):

Sweden has now essentially completed that transition, having just achieved <5% smoking prevalence. Interestingly, another shift further down the continuum of harm seems to be underway from snus to nicotine pouches in Norway and Sweden (Murphy et al., 2026):

Australia: Analysis of wastewater shows a substantial shift away from combustible tobacco to noncombusted, all while total nicotine consumption remained stable or slightly declining (Wang et al., 2026). (I think Alex Wodak would describe this as the inevitable transition to a less harmful product despite the regulations.)

Korea: Exclusive cigarette smoking showed large declines as use of e-cigarettes and HTPs, in various combinations, increased — all while total nicotine use was stable or slightly declining (Lee et al. 2026)

Conclusion

There was a stable proportion of the US population who used nicotine products over the past 10 years, but underlying that was a large shift away from combustible cigarettes to e-cigarettes. These changes were much more pronounced in younger age groups, while older adults mostly stuck with the same behavior (cigarette smoking with little uptake of e-cigarettes). Reassuringly, despite all the concerns about dual use, most e-cigarette use was exclusive use; dual use did not grow more quickly than exclusive e-cigarette use as critics claimed.

More broadly, this population shift down the continuum of harm is not unique to the US. Norway and Sweden showed similar shifts to snus, which displaced smoked tobacco rather than adding to it; the transitions in these countries are essentially complete. Other countries have at least the beginnings of similar transitions down the continuum of harm from cigarettes to e-cigarettes, HTPs, or nicotine pouches: a shift from one product to another, all while total nicotine use is stable or slightly declining.