Coolants in e-cigarettes are poorly researched: health impairments possible There is a lack of data, especially regarding possible effects on the lungs during inhalation

In the present opinion, a health risk assessment of substances in e-cigarettes or refill liquids (e-liquids) that produce a cold sensation (so-called cooling agents, hereinafter referred to as coolants) was carried out.

This assessment focuses on the coolants N,2,3-trimethyl-2-iso-propyl-butyramide (WS-23; CAS 51115-67-4; EC 256-974-4), N-ethyl-2-isopropyl-5-methylcyclohexane-carboxamide (WS-3; CAS 39711-79-0; EC 254-599-0) and ethyl 2-[[(1R,2S,5R)-5-methyl-2-propan-2-ylcyclohexanecarbonyl]amino]acetate (WS-5; CAS 68489-14-5; EC 695-735-2). The assessment was prompted, among other things, by the widespread use of coolants in e-liquids as well as by a report from Baden-Württemberg of a case of serious and ultimately fatal lung damage after consuming an e-cigarette containing WS-23. The concentrations of WS-23 in products on the market vary greatly in some cases. The responsible authority (Chemisches und Veterinäruntersuchungsamt Ostwestfalen-Lippe, CVUA-OWL) reports WS-23 concentrations between 0.55 and 0.86% by weight in analysed disposable e-cigarettes . In the EU Common Entry Gate (EU-CEG), concentrations of up to 4% by weight have been reported by the importer.

Opinion No. 045/2015 of the German Federal Institute for Risk Assessment (BfRshort forGerman Federal Institute for Risk Assessment), issued on 30 July 2015, assesses the health risks of additives in tobacco products and e-cigarettes, including WS-23, WS-5 and WS-3. Therein, the BfRshort forGerman Federal Institute for Risk Assessment recommends a comprehensive risk assessment for inhaled coolants and, in the context of the cooling and local anaesthetic effects (which facilitate inhalation of tobacco smoke), suggests that groups of substances with TRPM8 receptor agonists, such as substituted p-menthane compounds, should be examined and possibly banned, as they are expected to facilitate inhalation in a similar way to menthol. This has also been done for conventional smoking tobacco in the form of Annex 1 of the German Tobacco Products Ordinance (TabakerzV) due to the inhalation-facilitating properties of these substances.

The coolants WS-23, WS-3 and WS-5 are widely used in e-cigarettes and give consumers a cooling sensation which is often perceived as pleasant. Originally, these compounds were developed as part of a research programme to find refreshing substances for shaving products. Today, they are used in various products, including e-liquids, in which they simulate a cooling effect without affecting flavour. They act on the sensory nervous system, specifically on the TRPM8 receptor (Transient Receptor Potential Melastatin 8), which is responsible for the perception of cold. This receptor is normally activated by low temperatures, but also by agonists such as menthol, which signals cooling to the brain. Cooling agents such as WS-23, WS-3 and WS-5 are also TRPM8 receptor agonists and produce this effect without actually lowering temperature.

Although coolants such as WS-23, WS-3 and WS-5 are already used in various products, there is insufficient data on the inhalation toxicity of these substances, particularly in the context of e-cigarettes. Previous toxicological assessments of these substances are mainly based on oral or dermal exposureClose dialogexposureExposureTo glossary scenarios. As e-cigarettes allow direct inhalation of aerosols, respiratory exposure is of particular importance. There is thus an urgent need for specific data on inhalation toxicity and possible long-term health effects.

In this risk assessment for the coolants WS-23, WS-3 and WS-5, the Margin of Exposure (MOEshort forMargin of Exposure) approach was used and a Derived No Effect Level (DNELshort forDerived No Effect Level) was determined. Based on the DNELshort forDerived No Effect Level, a risk characterisation ratio (RCR) was performed to assess the health risk of inhalation exposure. For each of the substances mentioned, the dose at which no toxicological effects are observed (No Observed Adverse Effect Level, NOAEL) was derived from relevant toxicity studies and compared with the actual exposure. The resulting MOEshort forMargin of Exposure values show whether the current intake of the substance is likely to cause impairment to health. A low MOEshort forMargin of Exposure value or falling below a predetermined MOEshort forMargin of Exposure indicates an increased health risk. For substances that are neither genotoxic nor carcinogenic, the minimum value of an MOEshort forMargin of Exposure is usually 100 or above. If this value is not reached, a hazard to human health cannot be ruled out (EFSAshort forEuropean Food Safety Authority, 2023). In the present risk assessment, a safety margin of 200 was used for WS-23 and WS-5 in order to take into account the uncertaintyClose dialoguncertaintyUncertaintyTo glossary factor of 2 for the transfer of results from a subchronic study to chronic exposure and thus better cover long-term effects. For WS-3, an uncertainty factor of 6 was applied (= MOEshort forMargin of Exposure of 600) to extrapolate from subacute exposure (28 days) to chronic exposure (EFSAshort forEuropean Food Safety Authority, 2012a; ECHAshort forEuropean Chemicals Agency, 2012).

The MOEshort forMargin of Exposure is not a health-based guidance value, i.e. it is not a safety threshold below which daily intake is considered safe. Rather, the MOEshort forMargin of Exposure is used when there are indications of harmful effects, but the available information is not sufficient to derive what amount of the substance can be ingested daily without causing effects on human health.

For the sake of comparison, the DNELshort forDerived No Effect Level was also calculated in accordance with REACH Regulation (EC) 1907/2006. The DNELshort forDerived No Effect Level represents the derived exposure level below which the substance – taking into account all known uncertainties – leads to no impairment of human health. Furthermore, for risk characterisation, the RCR was calculated as the ratio between the estimated exposure and the DNELshort forDerived No Effect Level. If the estimated exposure is above the DNELshort forDerived No Effect Level (RCR ≥ 1), this indicates that a health risk is present (ECHAshort forEuropean Chemicals Agency, 2016).

Overall, the general data available on the coolants WS-23, WS-3 and WS-5 is very limited. In particular for the inhalation toxicity of these substances, there are no relevant, valid data. However, the available information on systemic toxicity after repeated oral administration indicates a potential health risk. As there are no relevant data on inhalation intake, 100% systemic availability is assumed. This assumption is based on a worst-case scenario due to the existing data gaps. EFSAshort forEuropean Food Safety Authority reports hepatic and renal toxicity from 10 mgshort formilligram/kgshort forkilogram bw (NOAEL: 5 mgshort formilligram/kgshort forkilogram bw/day) for WS-23, mild hepatic and renal toxicity from 40 mgshort formilligram/kgshort forkilogram bw (NOAEL: 8 mgshort formilligram/kgshort forkilogram bw/day) for WS-3, and renal changes, cardiomyopathy in female rats from 675 mgshort formilligram/kgshort forkilogram bw and haematological effects from 225 mgshort formilligram/kgshort forkilogram bw (NOAEL: 75 mgshort formilligram/kgshort forkilogram bw/day) for WS-5.

The evaluation of the available data shows that the reference MOEshort forMargin of Exposure is not reached and the RCR of 1 is exceeded for the three substances under consideration, WS-23, WS-3 and WS-5, in all exposure scenarios analysed, thus indicating a health risk for users. A comprehensive exposure assessment for these substances suggests that their inhalation via e-cigarettes exceeds exposure levels below which no health effects would be expected based on animal experiments. In particular, regular consumption of e-liquids with high concentrations of WS-23, WS-3 and WS-5 are expected to pose a risk to human health. For this reason, the use of e-liquids containing WS-23, WS-3 and WS-5 is not recommended from a toxicological point of view.

Studies have shown that synthetic coolants such as WS-23 can significantly improve the sensory experience when consuming e-cigarettes and increase the attractiveness of these products (Tackett et al.short foret alii (lat. "and others") 2023). These compounds exert their effect via TRPM8 receptors, which are responsible for the perception of cooling effects. Synthetic coolants can also have analgesic and anti-irritant properties. This effect can lead to increased consumption of these coolants. Higher concentrations of coolants in e-liquids in turn lead to increased exposure, which – especially with repeated use – leads to a health risk.

In addition, the apparent cooling effects of WS-23, WS-5 and WS-3 can also lead to an increased intake of nicotine, a change in inhalation behaviour, and possibly to increased dependence, especially among young and inexperienced users.

A comprehensive toxicological assessment of coolants in general is not possible due to the structural differences between the individual substances, so no generally valid statements can be made for all coolants currently available on the market.

The assessment is available in the PDF-Document.