Erna J Z Krüsemann1, Jeroen L A Pennings2, Johannes W J M Cremers3, Frank Bakker4, Sanne Boesveldt5, Reinskje Talhout6. 1. Centre for Health Protection, National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, the Netherlands; Division of Human Nutrition and Health, Wageningen University, Stippeneng 4, 6708 WE Wageningen, the Netherlands. Electronic address: erna.krusemann@rivm.nl. 2. Centre for Health Protection, National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, the Netherlands. Electronic address: jeroen.pennings@rivm.nl. 3. Centre for Health Protection, National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, the Netherlands. Electronic address: hans.cremers@rivm.nl. 4. Centre for Health Protection, National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, the Netherlands. Electronic address: frank.bakker@rivm.nl. 5. Division of Human Nutrition and Health, Wageningen University, Stippeneng 4, 6708 WE Wageningen, the Netherlands. Electronic address: sanne.boesveldt@wur.nl. 6. Centre for Health Protection, National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721 MA Bilthoven, the Netherlands. Electronic address: reinskje.talhout@rivm.nl.
Abstract
OBJECTIVES: Electronic cigarette refill solutions (e-liquids) are available in various flavor descriptions that can be categorized as fruit, tobacco, and more. Flavors increase sensory appeal, thereby stimulating e-cigarette use, and flavoring ingredients can contribute to e-cigarette toxicity. We aim to inform toxicologists, sensory scientists, and regulators by determining flavoring compounds in e-liquids with various flavors, and compare results between flavor categories. METHODS: Gas chromatography - mass spectrometry (GC-MS) was used to identify 79 flavorings in 320 e-liquids, classified in 15 flavor categories. Ten flavorings highly prevalent in e-liquids according to information from manufacturers were quantified. Flavoring prevalence was defined as the number of e-liquids with the flavoring as percentage of the total number of e-liquids. The method was validated in terms of specificity, linearity, repeatability, recovery, and sensitivity. RESULTS: The mean number of flavorings per e-liquid was 6 ± 4. Flavoring prevalence was highest for vanillin (creamy/vanilla flavor), ethyl butyrate (ethereal/fruity), and cis-3-hexenol (fresh/green). Based on similarities in flavoring prevalence, four clusters of categories were distinguished: (1) fruit, candy, alcohol, beverages; (2) dessert, coffee/tea, nuts, sweets; (3) menthol/mint; and (4) spices, tobacco, and unflavored. Categories from cluster 4 generally had less flavorings per e-liquid than fruit, candy, alcohol, beverages (cluster 1) and dessert (cluster 2) (p < 0.05). Flavoring concentrations varied between e-liquids within the categories. CONCLUSIONS: We evaluated flavoring compositions of 320 e-liquids using a simple GC-MS method. Flavoring prevalence was similar within four clusters of typically fresh/sweet, warm/sweet, fresh/cooling, and non-sweet flavor categories. To compare flavoring concentrations between individual flavor categories, additional research is needed.
OBJECTIVES: Electronic cigarette refill solutions (e-liquids) are available in various flavor descriptions that can be categorized as fruit, tobacco, and more. Flavors increase sensory appeal, thereby stimulating e-cigarette use, and flavoring ingredients can contribute to e-cigarette toxicity. We aim to inform toxicologists, sensory scientists, and regulators by determining flavoring compounds in e-liquids with various flavors, and compare results between flavor categories. METHODS: Gas chromatography - mass spectrometry (GC-MS) was used to identify 79 flavorings in 320 e-liquids, classified in 15 flavor categories. Ten flavorings highly prevalent in e-liquids according to information from manufacturers were quantified. Flavoring prevalence was defined as the number of e-liquids with the flavoring as percentage of the total number of e-liquids. The method was validated in terms of specificity, linearity, repeatability, recovery, and sensitivity. RESULTS: The mean number of flavorings per e-liquid was 6 ± 4. Flavoring prevalence was highest for vanillin (creamy/vanilla flavor), ethyl butyrate (ethereal/fruity), and cis-3-hexenol (fresh/green). Based on similarities in flavoring prevalence, four clusters of categories were distinguished: (1) fruit, candy, alcohol, beverages; (2) dessert, coffee/tea, nuts, sweets; (3) menthol/mint; and (4) spices, tobacco, and unflavored. Categories from cluster 4 generally had less flavorings per e-liquid than fruit, candy, alcohol, beverages (cluster 1) and dessert (cluster 2) (p < 0.05). Flavoring concentrations varied between e-liquids within the categories. CONCLUSIONS: We evaluated flavoring compositions of 320 e-liquids using a simple GC-MS method. Flavoring prevalence was similar within four clusters of typically fresh/sweet, warm/sweet, fresh/cooling, and non-sweet flavor categories. To compare flavoring concentrations between individual flavor categories, additional research is needed.