EC manufacturers have about 16,000 flavor chemicals from which to choose15. Our data provide a simpler picture: (1) the number that were used in our sample of 277 refill fluids was 155, not thousands; (2) in any given product, the number of flavor chemicals typically ranged from 0 to 50; and (3) while some constituents were present at rather low concentrations, 11 were found at concentrations >10 mg/ml. When evaluating just those chemicals that were over 1 mg/ml, the number per product ranged from 0 to 10. Moreover, the total concentrations of flavor chemicals exceeded the nicotine concentration in over half of the products. These data demonstrate that flavor chemicals are a major component of currently marketed EC refill fluids and their health effects on EC users should be addressed.
Of particular importance in our study is the finding that some products have individual flavor chemicals in concentrations >10 mg/ml, and many of these chemicals were found in many of the samples (e.g., ethyl maltol was in 24.5% of the products at ≥10 mg/ml, and menthol was in 22.6% of the products at ≥10 mg/ml) (Fig. 4). Based on the results of the MTT assay, menthol and ethyl maltol were present at concentrations that would be cytotoxic in 34% (26 of 76) and 40% (66 of 164) of the refill fluids that contained menthol and ethyl maltol, respectively. While the MTT data cannot be translated directly to in vivo human effects, they do raise concern about the potential for these chemicals to cause harm to users at the concentrations currently used in some refill fluids. Moreover, chronic exposure to high concentrations of flavor chemicals may be far more damaging than the effects seen in our acute experiments.
Further evidence that the concentrations of some flavor chemicals used in EC refill fluids may exceed safe levels can be found by comparing our data to the concentrations in other consumer products. Although cinnamaldehyde has been approved by the Food and Drug Administration (21CFR182.60) for use as a flavoring agent22 and given FEMA GRAS status, some in the flavor industry and the Research Institute for Fragrance Materials have recommended that cinnamaldehyde not exceed 1% when used in skin cosmetic products23,24. Cinnamaldehyde is usually found in body care and household products, such as detergents, creams and lotions, soaps and perfumes, in the range 0.001–0.8%25. Moreover, cinnamaldehyde is used in food products at concentrations ranging from 7.7 ppm (0.00077%) in ice creams to a 700 ppm (0.07%) in candy and up to a 6,400 ppm (0.64%) in fruits and juices23,26,27. In our refill fluid samples, two products had cinnamaldehyde concentrations of 118 mg/ml (11.8% or 118,000 ppm) and 343 mg/ml (34.3% or 343,000 ppm). We have previously reported that the cinnamaldehyde concentrations in a different set of refill fluid samples often exceeded 1% (range = 0.00022–14%) for cinnamon flavored refill fluids10,11. Our current study further shows, in agreement with our earlier work11 that cinnamaldehyde is more widely used in EC refill fluids than would be expected based on the names of the EC products. For example, cinnamaldehyde was found previously in fruity flavors, such as a product named “Blueberry Hills”, and in the current study was found in 70 of 277 (25%) products, even though only two products indicated “cinnamon” in their name. Cinnamaldehyde at concentrations found in EC products has also been shown to impair the function of immune cells in the respiratory system13.
Like cinnamaldehyde, ethyl maltol is added to edible products such as beverages, ice cream, candy, baked goods, gelatin desserts, meat, chewing gum and related products in concentrations up to 0.0142%28, and the maximum concentrations of ethyl maltol in final formulations of soap, detergents, and creams and lotions are 0.06%, 0.006%, and 0.01%, respectively29. These concentrations of ethyl maltol in consumer products are far below the concentrations that we found (0.008–3.13%) in 46% of the of the refill fluids that we tested. Ethyl maltol increases free radical formation in EC aerosols30, which could further increase the toxicity of products with this flavor chemical.
Menthol is commonly used in consumer products including tobacco cigarettes. Mentholated cigarettes generally have menthol concentrations <7 mg/cigarette and many are <0.002 mg/cigarette31. Menthol was present in 76 of our refill fluids at concentrations ranging from 0.002 to 68 mg/ml. Twelve out of the 76 refill fluids had concentrations greater than 10 mg/ml, which would exceed the concentrations normally found in conventional tobacco cigarettes flavored with menthol. Menthol produced cytotoxicity in the MTT assay at concentrations 30 times lower than the highest concentration found in the refill fluids we analyzed.
2,3-butanedione (diacetyl), which can cause bronchiolitis obliterans, also called “popcorn lung disease”32,33,34,35, has previously been found in EC products36,37. We found diacetyl, as well as two related chemicals, acetoin and 2,3-pentanedione, in 54% of the refill fluids. Of these chemicals, diacetyl, acetoin and 2,3-pentanedione were present in 36% (54 of 150), 42% (63 of 150) and 22% (33 of 150), respectively. Assuming a consumer vapes 3.4 ml of a refill fluid38 containing diacetyl at 0.32 mg/ml (highest concentration found in our study) and the transfer rate of diacetyl to the aerosol is 100%, the consumer would be exposed to 1.088 mg of diacetyl/day (equivalent to 85.83 ppb/8 hour average) which is well above the exposure limit of 5 ppb for 8 hours recommended by NIOSH39. Concentrations in refill fluids also exceeded the Short-Term (15 minute) Exposure Limit of 25 ppb for diacetyl39. These data raise concern about the potential for harm of some of the flavor chemicals that are present in refill fluids at relatively low concentrations.
Coumarin (1,2-benzopyrone) is another chemical of concern. It was present in 21 products at concentrations ranging from 0.007 to 5 mg/ml. Coumarin is currently prohibited as an additive in human food by the Food and Drug Administration (21CFR189.130) due to its hepatotoxicity, and when present, the food is deemed adulterated40. It’s prohibition in food supports the idea that it should likewise not be used in tobacco products, including ECs. Coumarin is often co-extracted from cinnamon with cinnamaldehyde and may have been a co-constituent inadvertently introduced into the products containing high concentrations of cinnamaldehyde
Our data show that both menthol and ethyl maltol are frequently used in refill fluids at concentrations that were cytotoxic to cultured human lung cells when tested with the MTT assay. Menthol and ethyl maltol have been reported in other brands of EC products41,42,43, although their concentrations were not given. While most prior work on the toxicity of EC flavors has been done on intact fluids9,44,45,46, several studies have examined the cytotoxicity of authentic standards of flavor chemicals present in EC fluids and aerosols41,47,48.
Our cytotoxicity data with menthol and ethyl maltol can be compared to results reported previously. Both ethyl maltol and menthol altered calcium homeostasis in CALU3 lung epithelial cells by depleting the endoplasmic reticulum of Ca2+ and elevating cytosolic Ca2+ 41. The effective concentration (EC50) of menthol in the Ca2+ assay (3.02 mM)41 was similar to the inhibitory concentration (IC50) of menthol in our MTT assay (1.38 mg/ml or 8.8 mM). In contrast, the concentration of ethyl maltol (0.15 mg/ml or 1.07 mM) that produced an effect in our MTT assay was much lower than the effective concentration (21.14 mM) in the Ca2+ influx assay41. These differences with ethyl maltol could be related to the different cell types (BEAS-2B versus CALU3) that were used in the two studies. These data show that mitochondrial reductase activity (MTT assay) is very sensitive to ethyl maltol and demonstrate the importance of evaluating multiple toxicity endpoints.
Cinnamaldehyde, which was very high in concentration in several products in the current study, was shown previously to be highly cytotoxic and immunosuppressive when tested in vitro with lung cells10,11,13,48. Based on our prior data with the MTT assay10,11, cinnamaldehyde is the most potent flavor chemicals we have tested, and it was found in 25% of all refill fluids in the current study.
Aerosolization of flavor chemicals can increase aldehyde concentrations in EC aerosols49, although this was not confirmed in a second study50. A previous study which compared the toxicity of EC aerosol produced at 3 versus 5 volts (4.3 W versus 11.9 W) showed a clear increase in toxicity at the higher voltage48. This observation would be consistent with the production of toxic reaction products upon aerosolization at the higher voltage and deserves further evaluation given the high concentration of flavor chemicals that we report here in many refill fluids, and the increased popularity of tank style EC with variable power controls.
