How do people actually transition between different nicotine products over time? Are young people who experiment with e-cigarettes more likely to become smokers, and is dual use a long-term pattern or simply part of the journey away from cigarettes?
Drawing on a decade of data from the U.S. Population Assessment of Tobacco and Health (PATH) Study, this presentation explores how nicotine product use changes among both underage and adult populations. Using longitudinal evidence that follows the same individuals over time, it examines real-world transitions between cigarettes, smoke-free products, dual use, and non-use.
Transcription:
GFN Science Lab 2 — Adam Lenart
2026-06-04 · 14:00-15:30 (Europe/Warsaw) · Global Forum on Nicotine 2026
Adam Lenart 0:00:11
Hi, everyone. So as Carrie mentioned, I will take on a bit of an ambitious task. Now in the next 10 minutes, I will make the other way. There you go. So I will touch on these three topics that each of them could be possibly a topic of one GFN. But it will be one slide each. So I will familiarize you with how we can see.
Thank you. So I will familiarize you a little bit with what is the PATH dataset, Population Assessment of Tobacco and Health. What kind of evidence can we see about product use in the United States from this dataset? Then I will make two bold claims. One of them is that the evidence of e-cigarette being a gateway to smoking is very limited in this dataset.
And another bold claim is that dual use is a transitory state that is expected and should not be something that is feared. But this is the way how a smoker, an exclusive smoker, in the end becomes an exclusive smoke-free products user. Again, take it with a pinch of salt, but I will guide you through in the next couple of minutes.
So, about the data set and why do we use this data set. So if you would like to understand product-use transitions, what we would like to have is to have a longitudinal data set so that we can measure a person across time. In the PATH data set, it is such a data set. It is a large US national representative data set of 50,000 plus consumers who are followed up across time.
Right now, we are at wave eight, the number eight follow-up, which has started in 2013. In this presentation, I'm going to be looking at the 2013 to 2023, so 10 years of time of product use experience.
And this is important because most of the literature that looks at, for example, gateways, looks at product use, looks at dual use, is not based on longitudinal data, but used on cross-sectional data.
And if, in a cross-sectional data, we see that, say, 10% of people are dual users, one year later, 10% of people are dual users, two years later, 10% of people are dual users, the observation is that dual use is constant. But if we can actually measure or follow the people over time, we can look at, are these the same 10% of people?
Or maybe more other people are transitioning into dual use while others are transitioning out. And I'm going to be showing you how this works. All right. My framework is going to be very simplistic. So in the PATH data set, again, it's a huge data set. It's 50,000 people and thousands of variables are measured in each of them.
So they are asked about what kind of product you use, what flavour, what brand, what nicotine concentration, is there peer pressure for you to use it, what's your religion, these type of questions, and even biomarkers. But so what I'm going to be using is something very simplistic. I'm going to be breaking down product use into four categories.
There are combustible category, cigarettes, cigars, pipes. There is smoke-free products category, which is mainly cigarettes. We are talking about US 2013, 2023. But there is also some smokeless tobacco, chewing tobacco in it. But again, it's primarily electronic cigarettes. We have dual use, which is using daily established use of combustible and smoke-free products.
And then we have cessation, which is not using any nicotine-containing products. Right. So again, I'm going to be using this longitudinal data set. I'm going to be extracting this product use data across time.
And I'm going to be measuring it in one part on adult people, but to address the youth question or the gateway question, I'm going to be linking the path youth data set, underage data set with the adult data set and following them up how the youth, younger generation who becomes adults are followed up in the adult sample.
So let's look at what we see from adult life course transitions across time. Here I'm going to be looking at two transitions. One transition is wave one, so how people from 2013 to 2015 behaved. What you see in this diagram is what happens to a hypothetical or typical 18-year-old smoker. An 18-year-old smoker has a 100% chance of being an 18-year-old smoker, right?
But over time, for example, by age of 75. If you'd like to hear more about the methodology, please check the video or ask me after this presentation. But I'm using basically the same methodology as what demographers use for calculating life expectancy at birth. So what we see is what we expect, that about roughly 30% of smokers would be lifetime smokers.
And about by age 50, 50% of smokers would have switched out of smoking. So what we see here in the combustible part of the diagram is the probability of being a smoker at some specific age.
And then what we see is that in orange, the probability of switching to dual use with this little sliver of green is the probability of switching to smoke-free product use and the gray area is the probability of switching out or cessation. So if we juxtapose the 2013 data with the data eight, 10 years later, what we can see is that transitioning out of smoking has sped up.
People spend less time smoking. What do they do instead? They can be dual users, which also reduced a bit compared to the 2013. What has increased a lot is the smoke-free product use. So how behaviour has changed in these 10 years is that there are more and more smokers who switch to smoke-free products. And I will claim later that this happens through a period of dual use.
But what hasn't changed, although there is now a new product category on the market, is the cessation. There is still the same number of people who transition out of any nicotine-containing products.
So what we can see is that smoke-free products, rather than crowding out cessation, they rather show an additional pathway that people can decide how they want to consume their nicotine.
More specifically, Just a few numbers of what happens in 2021, 2023, is that a smoker, like a hypothetical or typical 18-year-old smoker, by age 35 or so, 35% would have quit, about 5% would have transitioned to dual use, and 9% directly to smoke-free products.
So again, PATH is an extremely rich data set that you can draw some conclusions about long-term product use due to the granularity and the longitudinal follow-up of the data. So now moving on using a similar diagram. So that was a typical 18-year-old smoker and following their decisions across their life course. So what happens with a typical 18-year-old users?
What we see, obviously, even at 2013, that there is basically nobody who stays a dual user for their whole life. There is a very, very low chance for that happening. This is even more apparent in the latest wave. Well, 1%, 2% of people are expected to stay dual users across their life course. So in the first wave, again, how consumer behaviour has changed.
In the first wave, if you were not a dual user, your first primary decision or most likely decision at younger adulthood was to switch back to smoking. The pink area is the fattest in the beginning, while cessation picks up later in the life course. And some dual users have transitioned, of course, to smoke-free products. And again, let's compare it with the later waves.
Dual use period has shortened, or is expected to shorten. A dual user, again, this typical 18-year-old dual user has roughly a 25% chance to stay dual users in 10 years. So 75% of them would have done something different. And most of them would have quit. or move to a smoke-free product, while a smaller portion of them would move back to exclusive cigarette smoking.
I will look at the animation in two minutes, how this works in a dynamic setting. But if we have a group of combustible users and we look at the transition probabilities, like what is the likelihood that from one year to another, you stay a smoker. If you stay a smoker, that's your probability. If you move to a dual user, we have this small pink line.
If you, a combustible user may directly move to a smoke-free products user or a combustible user may directly move to a, become a non-current users. And the width of each of these arrows describes the probability. So what we can see is that combustible smokers most likely stay combustible smokers. But if they switch to dual use, we have a much higher probability.
It's actually 1.2% likelihood of a combustible smoker moving to dual use. smoke-free products and 9.6% for a dual user and your probability into becoming an exclusive smoke-free products user over the whole study. But again, I will show you how this works in a minute. Now, oh, and again, what does this all mean?
that, of course, if you see a survival function like this, it means that people don't stay dual users for very long. If they don't stay dual users for very long, this indicates that, of course, it's not an expectation of something constant across the life course, but it's a really dynamic state that leads to either switching to smoke-free products or cessation.
Okay, that was claim number one, dual uses being transitory and not a stable category. Claim number two, gateway. So, of course, one big issue or fear of electronic cigarettes that new initiators of electronic cigarettes would become smokers eventually.
So from the PATH data, we can also get a glimpse at that claim and have just some simple descriptive statistic to look at this. And again, it's a simple analysis, just looking at people, what they do and when they do it. But because of the richness and the representative of the data, this can allow us to take these findings a bit more seriously. So we have two plots.
One of them is e-cigarette initiators who have never smoked before. Sorry, underage e-cigarette initiators who have never smoked before. And we have underage combustible initiators who have never vaped before, right? We have the smokers and we have the e-cigarette users. These are our two groups. We have 255 people, underage e-cigarette users. e-cigarette initiators.
And we have 41 in the PATH data set. Again, 41 is not a big number, but it's not a big number in a data set of multiple thousands because there is not many of them in the United States. And so in the first column, we have 100% of these underage initiators being underage initiators. And we are following them up, what they do in wave six.
So, you know, one and a half, two years later, until Wave 7. And what we see is that these e-cigarette initiators, two-thirds of them, stop smoking and not taking any tobacco or nicotine anymore. Again, this is... You know, it's also similar to what we have seen before. Biggest changes are in the youngest ages, and most nicotine initiators will not use nicotine.
Roughly one-third of them will keep using e-cigarettes, and there is just a little, little percent of them, 0.4, so that's one person out of 255 who have become an exclusive cigarette user, and three 1% of them who were using e-cigarettes and cigarettes. Again, just to make sure, let's follow them up one wave later.
So here they are underage, here they are young adults, and here they are two years older, younger adults, so like 18, 20, 22. sorry, 17, 19, 21. And at the end, we have roughly 1% of them being exclusive combustible tobacco user and 2% dual using, so roughly 3% of them being tobacco users.
which is, again, as a naive statistician and not a behavioural scientist, this indicates to me that I, as a tobacco harm reduction scientist, I don't really need to be worried about this.
Because if we were to see 100% of e-cigarette users and 80% of them moving to tobacco, then this would indicate, yes, there is a strong gateway effect, but we don't see such a huge transition. Again, it's out of 100% of them, 3% would consume tobacco. OK, let's juxtapose it with what the actual tobacco smokers do. We have this small number, 41 smokers.
Again, roughly 60, about 60%, two-thirds of them will not use tobacco anymore. About 20 plus 7, 27% of them would still smoke combustible cigarettes or pipes or cigars. And about 10% of them would have switched to electronic cigarettes.
So if there is any gateway, it's more like smokers moving to e-cigarette use rather than the other way around that e-cigarette users moving to combustible cigarettes. But again, we are talking about United States where cigarettes are socially not that accepted as in other countries. So again, these results pertain to United States about 2020.
OK, so my last slide before the conclusions is going to be showing how a starting cohort of combustible smokers. You will see a big red or purple blotch of points appearing here very soon.
And what we will do is that we start in 2013, wave number one, and we'll show you how these individuals, there will be these circles that you will appear, each circle is one individual, and how these individuals will move from different categories, being a combustible smoker to a dual user, an exclusive smoke-free products user, or a non-current, which is not using any tobacco or nicotine products anymore user.
And we are going to follow up what is the ratio of smoke-free product users who transition from combustibles through dual use and not directly from combustibles to smoke-free products. Could we start the animation, please? OK, so as I promised, the big blob of smokers appeared. We are wave one. Now we have just selected the combustible smokers.
And two years later, just 1% of them switched to smoke-free products, and about 6% of them are dual users. And as time goes by, what we can see is that, Most combustible smokers, as I said before, become non-current users, which are not using any products anymore.
Dual use is beefing up as the combustible smokers start to experiment with electronic cigarettes, while some of the smoke-free products users with the orange dots, we can see those who have transitions from dual use to smoke-free products use.
And with the blue dots, we can see those who have transitioned to smoke-free products, either from being combustible or a non-current user. Now we are at Wave 7. I don't know if we can go to the end.
Well, if you have a good memory, what you could see is that we had our final blob of electronic cigarette exclusive smoke-free products users, which are mostly electronic cigarette users in the United States. And most of them became an electronic cigarette user through a period of dual use during this time. again, all of them were smokers.
More than half of them switched to exclusive smoke-free products used through a transition of dual use. And what does this tell us? That it takes time to switch to exclusive smoke-free products. I mean, these products are technological innovation. It's not a magic pill that we can drop one day.
Some people choose to do so, but many people, as you would expect, for example, me as a naive statistician, if I see a time series of 20 cigarettes a day, 20 cigarettes a day, 20 cigarettes a day, my expectation is going to be 20 cigarettes a day. If it's going to be 15 cigarettes a day, I'm going to celebrate it. But I don't expect that it's going to be zero cigarettes a day.
Right away, over time, if people decide to do so, and they have shown the willingness to reduce their cigarette smoking, this does reduce. And this is exactly what we see in the data. There are people who decide to become exclusive smoke-free products users right away. But there are people who simply don't. Maybe it doesn't fit their lifestyle.
Maybe it doesn't fit their taste right away. But over the years, it takes time. And many of them do decide to become exclusive smoke-free products users.
So what I have tried to show you is that we actually do need this longitudinal complex data sets to really understand this transition dynamics, because these are not simply linear process that here is an electronic cigarette, take it, and then people magically become electronic cigarette users. People have their habits. They have the choices they want to make.
And if you can follow them up with a longitudinal data set, we can really understand what matters to them and how they want to consume their products. What we have also seen that dual use is not a constant state, but this is a transitory state. that which is part of the consumer journey from being a cigarette smoker and becoming an exclusive smoke-free products user.
From the first analysis, we have also seen that in this data set during this timeframe, smoke-free products did not crowd out cessation just because a new product category was added to the was introduced to the US market, it did not mean that people stopped quitting.
What happened is that smoke-free products provided an additional pathway that crowded out smoking and sped up switching from smoking to a lower risk product. And when we looked at youth, we show a very minimal, about 3% chance that an electronic cigarette user became an exclusive smoker.
And of course, the question is, what would they have done if there were no electronic cigarettes on the market? So overall, what we can see that smoke-free product use in this data set is consistent of a narrative with a progressive displacement of smoking. Thank you so much.
Carrie Wade 0:23:24
Having three GFN Fives in one presentation. If there's any questions, we probably have time for one or two questions. Sorry, we probably have time for one or two questions. Well, I will ask my question, and thank you for not having your own. So we know that this data comes from the US, and I have two questions.
One is, at some point during those waves, there was the change to T21. Was that factored in? went into the underage use.
Adam Lenart 0:23:56
No, so the data set still follows up and says that 18, it still follows up from 18 euros, so the sampling frame hasn't changed.
Carrie Wade 0:24:05
Great, okay, that's one question I had. The other is, is there potential for the PATH study, because it is a US-focused study, to be applied to other countries or regions like the EU? I know this is desperately needed information, particularly with the TPD coming along, and all the iterations of the TPD that will happen in the future.
Adam Lenart 0:24:25
Yeah, so I think there is a huge potential and need, but in Europe currently what we mainly use is the Eurobarometer, which is a cross-sectional follow-up of, you know, two, three hundred users or maybe one thousand from bigger countries, which does not have the same granularity.
We had household panel surveys in the early 2000s, which had some smoking questions, but those were cancelled, I guess, because of budgetary reasons.
0:24:55
Yeah.
Adam Lenart 0:24:56
So I think PATH is a really good way of spending taxpayers' money. As I understand, it's mainly funded by the taxes levied from the tobacco industry.
Carrie Wade 0:25:05
Right.
Adam Lenart 0:25:05
Yeah.
Carrie Wade 0:25:05
So it's free for the Americans.
Claude Bamberger 0:25:08
Here we go. I've got a question on part of an answer for your question. Did you do the same logic, but going in reverse to analyse how people are coming there. Because in France, we had a study which was quite interesting, where they tried to elucidate if people go through vaping at a moment, will they become daily smokers?
And what they showed is a protective effect of going through vaping. It's like in the US, it's the most common a smokeless product, there was a 40% protective effect on becoming a daily smoker.
A bit the same thing you did, but they did it in reverse, analyzing how are they coming or not, because like you said, The big question is, what would they do if there wasn't this opportunity? And basically, in the population, they did that in high school. Basically, you have people that begin to smoke and continue to smoke and become daily smokers.
And you have people that begin to vape. or begin to smoke and then vape, and at the end, there is 40% less daily smokers in the true population.
Adam Lenart 0:26:49
Yeah, that's interesting. No, I haven't tried it, but my expectation is that it may be similar. If the revealed preference is already that you want to try a low-risk product, then it already indicates that these people would probably like to keep trying a low-risk product. Thank you.