Aldehyde emissions from e-cigarettes: replication studies challenging previous reports


By Dr Farsalinos

On December 15 2016, I was asked about my predictions for 2017 about the future of vaping and I mentioned that: “Bad quality studies accompanied by impressive press statements are becoming increasingly frequent … I think we will see some interesting developments in this aspect soon.

It took a few months but it is now time to deliver on that promise. You all remember the “hidden formaldehyde” research letter, which included a statement that e-cigarettes are 5 to 15 times more carcinogenic than smoking. The letter was accompanied by impressive media coverage through a press statement and a later award for the research group of $3.5 million for further research on this issue.

Few days ago, we published a replication study, using exactly the same e-cigarette devices, batteries and liquid as the research letter above. However, we asked vapers to try the device (a very old CE4 atomizer, tested in 0.2 V increments) and report dry puffs when detected. Dry puffs were detected at 4.2 V, so 4.0 V was the maximum realistic use voltage. We then tested the device in the laboratory at different voltage settings (both realistic use conditions and dry puffs) and we found, as expected, an exponential increase in formaldehyde emissions at dry puffs. In fact, at 5.0 V we found much higher levels of formaldehyde than the original research letter. But we did not even dare ask vapers to try that device at 5.0 V. So, we said that e-cigarettes can generate huge levels of formaldehyde, but this happens in conditions that no vaper will ever be exposed. We clarified that the 5 to 15-fold higher cancer risk is wrong. We also recommended that vapers should not use these devices (CE4s) because even in normal vaping conditions the level of formaldehyde emissions was quite high and much higher than recent devices. Of note, CE4s are not available in the European Union any more.


In 2016, another study found huge levels of aldehyde emissions from e-cigarettes. They used again a CE4 atomizer (CE4v2) at 3.8 V and 4.8 V but with 5 second puffs (puff duration is equally important to power or voltage because energy = power x time). They reported extreme emissions (up to 48,000 ug/g formaldehyde). We calculated the relative exposure from 5 mL liquid consumption (assumed daily use) compared to tobacco cigarettes and we found that 5 mL of liquid would be equivalent to 3200 tobacco cigarettes!! The authors expanded by publishing a second study, in which they performed a risk assessment analysis for vapers and passive exposure using the previous findings.

Today, we published a study replication using the same equipment, power settings and puff duration. Two experienced vapers identified dry puffs at both 3.8 V (obviously due to the very long puff duration) and at 4.8V (in fact, even at shorter puff duration, dry puffs were horrible at 4.8 V). This time, the problem was not only dry puffs; the reported results were also hugely overestimated. We found 6 to 25-fold lower levels of aldehydes at the same dry puff conditions. Furthermore, to assess aldehyde emissions from a realistic use pattern, we tested a relatively new atomizer (Nautilus Mini, which is in fact about 3-years old but has a cotton wick as all new-generation atomizers). The levels of aldehyde emissions were so low that a liquid consumption of 5 mL per day would expose vapers to 94.4-99.8% lower aldehyde levels compared to smoking 20 tobacco cigarettes. Please note, the comparison refers to 5 mL liquid consumption; the more you consume, the less the difference compared to smoking will be. The association is linear.


Both studies clearly show that it is highly important to evaluate for the generation of dry puffs when measuring e-cigarette emissions in the laboratory. Although this has been known for years (from vapers) and has been mentioned in the literature since 2013, still many (if not most) studies fail to examine this.

The field of e-cigarette research has an unusually high number of studies reporting “strange” (to say they least) results. A basic principle that should have been followed is that if the data do not make sense, don’t blame common sense but look at the data for possible mistakes. Still, many (if not most) of these studies are accompanied by press statements and widespread media campaigns. As a result, today, after so many years of research, the majority of smokers think that e-cigarettes are equally or more harmful than smoking. It would be interesting to see how the journals and editors who publish these studies will react when the findings of their publications cannot be replicated.

I would suggest everyone to stay tuned. This is NOT the end of replication studies. More is coming soon, as I promised in the past.



Background Image

Header Color


Content Color