- A new paper claims vape liquids can release harmful chemicals when overheated.
- Experts say the tests used in the study don’t match how people actually vape.
- Human studies show no measurable breathing or eye problems after short-term exposure to high levels of propylene glycol.
- Harm reduction scientists say some research keeps repeating old claims about harm without looking at real-world use.
A new study claiming common vape ingredients release harmful chemicals when heated has been strongly criticised by harm reduction experts. They say the findings come from unrealistic lab tests that don’t reflect how people actually vape.
The ‘mini review’, published in Toxicology Letters on 1 October 2025, is titled “Toxicity of humectants propylene glycol and vegetable glycerin in electronic nicotine delivery systems.” It examined studies on propylene glycol (PG) and vegetable glycerin (VG) – the two main liquids used in vapes – and suggests that when they are heated, they can produce toxic chemicals.
According to the authors, “toxicity associated with PG/VG inhalation is primarily due to thermal degradation byproducts (TDBs) generated by PG/VG-containing e-liquids.”
They added: “High-power ENDS devices with sub-ohm power capabilities generate aerosols with larger mass and higher concentrations of TDBs. The most common TDBs identified in e-cigarette aerosols include formaldehyde, acetaldehyde, acrolein, acetone, acetoin/diacetyl, as well as benzene.”
In plain terms, the paper argues that when vapes are run at very high power, they can create small amounts of some of the same chemicals found in cigarette smoke.
But harm reduction researchers say this conclusion is misleading because it comes from tests that don’t match real-world vaping conditions.
‘Harm-searching’ in vaping research
Researcher Robert Sussman said the review is another example of what he calls the “harm-searching literature” – studies that re-analyse old data to highlight potential risks.
“The harm searching literature is extensive and keeps growing, so it’s always possible to find the ‘right’ reference supporting your harm claim,” he said. “In this.. it was the turn of PG/VG.”
He said that while heating PG and VG can produce byproducts, “as long as the devices are puffed as intended by their design, the byproducts remain at trace levels.”
“High-powered sub-ohm devices must be puffed as intended by their design: with deep inhalation (direct to lung), otherwise the user gets a nasty taste of a hot aerosol,” he said.
Sussman explained that many of the studies cited in the review use unrealistic conditions. “Unfortunately, many emission studies test these devices under unrealistic conditions (puffed with the airflow that is used in pods). This produces excessive toxicity (but users would find these aerosols repellent). This is not a frivolous claim, it is supported by thermal physics.”
Human studies show no harm from short exposure
Sussman pointed to a 2018 study published in Inhalation Toxicology that looked at the effects of breathing PG for several hours.
Ten men and ten women were exposed to PG aerosols for four hours at levels of 20 and 100 milligrams per cubic metre, and for 30 minutes at 200 milligrams per cubic metre – far higher than people would experience while vaping. Participants exercised lightly during the test and were checked for eye irritation, dryness and lung function.
The results said: “Objective measures of pulmonary function and ocular irritation did not reveal any exposure-related changes. Exposure-related changes in symptom reporting were observed; however, the highest symptom ratings did not exceed ‘slight’ on the scale.”
The researchers concluded: “The results indicate at the concentrations and acute durations tested, PG does not affect human respiratory function or produce ocular irritation.”
Sussman said this shows that “PG does not affect human respiratory function or produce ocular irritation,” even at doses “about 100 times the PG dose inhaled while vaping.”
No evidence of serious harm after 15 years of vaping
Sussman said that after more than a decade of vaping, “there is no indication of any serious harm, perhaps some minor irritation or wheezing.” While he noted that long-term data beyond 30 or 40 years is not yet available, there is currently “no indication” of major health problems linked to PG or VG.
Both ingredients have been safely used for decades in foods, medicines and stage fog machines. Harm-reduction experts say that repeating extreme lab tests risks misleading the public and policymakers.
“When devices are overheated beyond normal use, you can generate all sorts of chemicals,” said one researcher. “But users wouldn’t inhale that vapour – they’d stop immediately because it tastes burnt.”
