- According to an online survey by Truth Initiative, 59.1% of 15- to 25-year-olds who vape in the United States had used a smart vape in the previous month.
- The strongest associations involve battery and e-liquid gauges, not video games or phone connectivity.
- These gauges are mainly found on large, high-nicotine disposables, a factor the study does not control for.
- Several results presented as robust do not survive the statistical correction the authors say they applied.
- In 2026, the FDA authorized connected pods with several of the features listed in the study, to prevent their use by minors.
An online survey of 4,719 young vapers
E-cigarettes have changed a great deal since they first appeared on the market in the mid-2000s, and some of these changes raise questions. This is particularly true of smart vapes, products with an attractive, often colorful design and a range of features alongside vaping. Some models, for example, offer video games or let users read their text messages.
The study discussed here was published on November 4, 2026, in the journal Tobacco Control. Authored by Jennifer M. Kreslake and colleagues at Truth Initiative, a US anti-tobacco organization, it looked at smart vapes and their impact on young people’s nicotine dependence.
The data came from an online panel surveyed between July and September 2025. The researchers included 4,719 young people aged 15 to 25 who had vaped on at least one day in the previous month. Those who had used a smart vape were asked which of twelve features their device had: notifications (low battery, charging complete, ready for use, low e-liquid, distance from the phone), location tracking and remote locking, settings (nicotine level, flavor intensity, coil temperature), video games, and on-screen gauges (battery, e-liquid, puff counter).
Each feature was then compared against two indicators: the number of days spent vaping in the month, and a dependence score based on four questions, such as cravings becoming intolerable after a few hours without nicotine, rated from 0 to 4.
For context, the authors rely on the 2024 National Youth Tobacco Survey (NYTS). Its 2025 edition, published in June 2026, reports current e-cigarette use among 5.2% of US middle and high school students, down over the 2022–2025 period.
Striking figures
In the sample, 59.1% of respondents had used a smart vape in the previous month, and 19.7% had already downloaded the associated app. The most common features were also the simplest: the battery gauge (35.5%), the e-liquid gauge (25.8%) and the puff counter (23.6%). Video games concerned only 8.6% of respondents.
These most common gauges were also the ones showing the strongest associations. Users of a device with an e-liquid gauge had roughly three times higher odds of reporting more frequent vaping, and those with a battery gauge 2.76 times higher odds. Nicotine level control came third (2.19). On dependence, the largest score differences concerned the phone-distance alert (+0.26 points) and the low battery notification (+0.18 points).
Based on these figures, the authors concluded that these digital and interactive features should be closely scrutinized in tobacco control policy and during the regulatory review of products.
The “addictive” features show almost nothing
The narrative around smart vapes usually rests on their resemblance to smartphones: games that reward puffs, notifications, constant connectivity, and so on. Yet, as the researchers’ own figures show, it is precisely these features that produce the weakest results.
Location tracking, remote locking and the “ready for use” notification are associated with neither frequency nor dependence. Video games, for their part, are associated with a slightly higher dependence score (+0.14 points), but not with more frequent vaping.
| Category | Feature | Odds ratio | 95% CI | Significant |
|---|---|---|---|---|
| Gauges | E-liquid level | 3.12 | 2.75–3.56 | Yes |
| Gauges | Battery level | 2.76 | 2.45–3.10 | Yes |
| Settings and games | Nicotine level | 2.19 | 1.90–2.54 | Yes |
| Gauges | Puff counter | 1.85 | 1.62–2.10 | Yes |
| Notifications | Low battery | 1.83 | 1.62–2.08 | Yes |
| Settings and games | Flavor intensity | 1.79 | 1.54–2.09 | Yes |
| Notifications | Low e-liquid | 1.67 | 1.44–1.93 | Yes |
| Notifications | Charging complete | 1.51 | 1.32–1.73 | Yes |
| Settings and games | Coil temperature | 1.33 | 1.10–1.60 | Yes |
| Location | Remote locking | 1.17 | 0.94–1.44 | No |
| Location | Device tracking | 1.07 | 0.90–1.28 | No |
| Settings and games | Video games | 1.04 | 0.85–1.27 | No |
| Notifications | Ready for use | 1.01 | 0.87–1.16 | No |
| Notifications | Distance from the phone | 0.83 | 0.70–1.00 | Borderline (less frequent use) |
| Source: Kreslake et al., Tobacco Control, 2026 (tables 3 to 5). Odds of reporting more frequent vaping in the past month, compared with users of devices without the feature (1 = no difference). Each feature was tested in a separate model, adjusted for age, sex, race/ethnicity, financial situation, use of other tobacco products and app download. | ||||
Stranger still, the phone-distance alert is associated with less frequent vaping but with higher dependence. The study does not comment on this contradiction.
Gauges that mainly point to a product type
A battery or e-liquid gauge is no gimmick: on a disposable rated for tens of thousands of puffs, the screen is first and foremost there to avoid unexpected cut-outs. The authors themselves acknowledge that these features are more utilitarian than games or customization options.
Above all, they are found on the giant disposables that dominate the US market. The authors themselves cite the Geek Bar Pulse and RAZ, which they say were among the five best-selling brands in the United States in August 2025.
In the US, these products are sold with nicotine concentrations of up to 50 mg/ml. Since nicotine is addictive, it seems logical that someone vaping an e-liquid at this strength would be more likely to show signs of dependence. Yet the nicotine strength of the products used was not taken into account in the calculations.
The very direction of the relationship the authors try to establish is also questionable: a heavy user has every reason to choose a device that shows how much charge is left. In other words, when the authors suggest these gauges could encourage use, the reverse is just as plausible: an already dependent user may have chosen a model with this gauge precisely so as never to run out.
The comparison group further muddies the picture. For each feature, users of the device in question are compared with a group that mixes vapers who do not use a smart vape and users of smart vapes that lack the feature. Each indicator therefore largely measures the simple fact of using a smart vape.
Statistics shakier than claimed
To account for the large number of tests (twelve features, two indicators), the authors also say they applied a Bonferroni correction, and claim that all their significant results survived it. Yet their own confidence intervals tell a different story.
A bit of scientific jargon
A Bonferroni correction is a statistical safeguard against false positives when researchers run many tests at once.
Take an example: imagine a six-sided die, and designate 6 as a suspicious number. If someone rolls the die once, the chances of getting a 6 are relatively low. If that person rolls the die 24 times, however, those chances rise dramatically. This is precisely what a Bonferroni correction seeks to avoid.
In this study, the researchers ran 24 tests. With so many tests, the probability that at least one result is significant by pure chance is about 71%, compared with 5% for a single test.
A Bonferroni correction therefore lowers the significance threshold of each test. Instead of the 5% margin of error usually accepted for a single test, the threshold drops to 0.2%: 0.05 divided by 24 (the number of tests) gives about 0.002, or 0.2%. With 24 tests each judged at 0.2%, the overall risk of getting at least one false positive returns to about 5%, the usually accepted margin of error.
According to our calculations based on the published figures, the association between the battery gauge and dependence corresponds to a p-value of about 0.03, that of video games to about 0.01, and those of nicotine and flavor settings to about 0.006–0.007. With 24 tests and a corrected threshold of about 0.002, none of these results falls below it. These results are therefore no longer significant once the correction is applied.
The dependence gaps linked to these features are also small. The score ranges from 0 to 4, and respondents score 2 points on average. Using a device with one of these features raises it by only 0.07 to 0.26 points, far less than the gap that usually separates two respondents picked at random.
Recommendations that overreach
In their study, the authors also argue that smart vapes run counter to Article 13 of the WHO Framework Convention on Tobacco Control (FCTC), which deals with advertising, promotion and sponsorship. The argument calls for two caveats.
First, the United States, where the survey was conducted, signed the convention in 2004 but has never ratified it. Second, the wording used by the authors, which mentions video games and mobile apps, comes from a decision of the Conference of the Parties adopted in 2024 on cross-border advertising and the depiction of tobacco in entertainment media. The Article 13 guidelines do consider that a product’s design can amount to promotion, but this passage targets the way content depicts or promotes tobacco, not the features built into a device. The argument is therefore more a matter of advocacy than a legal obligation for the United States.
Finally, the authors claim that the FDA has not authorized any smart vape, and even this point is debatable. In 2026, the agency authorized the Glas G2 pod, which must be paired with a smartphone via Bluetooth, stops working if separated from it and runs random biometric check-ins. On August 28, 2026, it authorized the Juul2, also linked to an app that can lock the device. Phone connectivity, locking, proximity-based operation: several of the features listed in the study are therefore found on authorized products. There, however, they serve to prevent use by minors.
What about France?
The smart vape debate is of little concern to France. These products are mostly disposable e-cigarettes, banned from sale since February 2025. The study was nevertheless picked up by Génération sans tabac, the website of the French National Committee Against Smoking (CNCT), under the headline Une plus forte dépendance chez les jeunes utilisateurs de fonctions de « smart vapes » (“Stronger dependence among young users of smart vape features”). The article notes that the study cannot establish cause and effect, but its headline presents as a finding what is only an association. And as usual, the image the association shared on social media mentions none of the study’s limitations.









