Nicotine, Humectants, and Tobacco-Specific Nitrosamines (TSNAs) in IQOS Heated Tobacco Products (HTPs): A Cross-Country Study

Abstract

Heated Tobacco Products (HTPs) purport to reduce exposure to tobacco-related toxicants compared to combustible cigarettes. This cross-sectional study examined the content of nicotine, two humectants (propylene glycol (PG) and vegetable glycerin (VG)), and four tobacco-specific nitrosamines (TSNAs: NNN, NNK, NAT, and NAB) in the tobacco filler of a popular HTP brand (IQOS). Non-menthol and menthol IQOS sticks were purchased from nine countries between 2017 and 2020 and were classified into two versions ("Bold" and "Light") using Philip Morris's flavor descriptors. The average nicotine concentration was 4.7 ± 0.5 mg/stick, and the highest nicotine concentration was found in products from Japan (5.1 ± 0.2 mg/stick). VG was the dominant humectant found in all sticks, with an average concentration of (31.5 ± 2.3 mg/stick). NNN, NNK, and NAT were substantially higher in the "Bold" sticks than the "Light" sticks. Significant differences between countries for TSNAs were also observed: the NAT and NAB contents were the highest in the "Light" products from Canada (192.5 ± 24.1 and 22.9 ± 1.0 ng/stick, respectively); the NNK concentration was the highest in the "Bold" products from Poland (64.8 ± 7.9 ng/stick); and the highest NNN concentrations were observed in the "Bold" products from South Africa (488.9 ± 26.7 ng/stick). As NNN and NNK are known human carcinogens, and as humectants like PG and VG can degrade into toxic carbonyl compounds upon heating, monitoring the concentration of these chemicals in HTPs is important for protecting users' health and ensuring compliance with regulations.

Keywords: IQOS; TSNA; heat-not-burn; heated tobacco products; humectants; nicotine.

Figure 1

“Bold” vs. “Light” IQOS stick classification. Classification of “Bold” vs. “Light” was completed using PMI’s flavor descriptors on its country-specific websites; see Methods. For products purchased in multiple countries, only one example per product type is presented above. Small package differences were observed between countries; all sticks used in this study had flavor descriptors on their country-specific websites. Packaging for IQOS White was damaged during experiments. Key: Canada (CA), Israel (IL), Italy (IT), Japan (JP), Poland (PL), the Republic of Korea (KR), South Africa (ZA), the United Kingdom (UK), and the United States (US).

Figure 2

Chemical characteristics of “Bold” and “Light” IQOS sticks purchased in non-menthol and menthol flavors; see Figure 1 for classification. Mann–Whitney non-parametric t-tests were used to compare each study variable between “Bold” and “Light” for non-menthol and menthol sticks. Error bars show standard deviation. Number of product types per category: non-menthol “Bold” n = 5, non-menthol “Light” n = 7, menthol “Bold” n = 3, menthol “Light” n = 7. All products were extracted and analyzed at least three times. Key: (a) weight, (b) nicotine, (c) vegetable glycerin (VG), (d) N’-nitrosonornicotine (NNN), (e) 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), (f) N’-nitrosoanatabine (NAT), and (g) N-nitrosoanabasine (NAB); * = p < 0.05 compared to “Light” within flavor (i.e., menthol or non-menthol); see Results for exact p values.

Figure 3

Chemical characteristics of “Light” IQOS sticks compared by country of purchase. Kruskal–Wallis non-parametric ANOVA with Dunn’s multiple-comparison tests was used to compare the mean rank of the country with the lowest concentration of each variable to the mean rank of each other country. Error bars show standard deviation. For comparisons of mean rank of each country with every other country, see Table S2. Number of “Light” products per country; CA n = 1, IL n = 2, IT n = 2, JP n = 1, PL n = 2, ZA n = 2, KR n = 1, UK n = 1, US n = 2; see Figure 1 legend for country abbreviations. All products were extracted and analyzed at least three times. * = statistically significant (p < 0.05) between the country with the lowest value and this country; see Results for exact p values. See Figure 2 legend for a key.

Figure 4

Chemical characteristics of “Bold” IQOS sticks compared by country of purchase. Kruskal–Wallis non-parametric ANOVA with Dunn’s multiple-comparison tests were used to compare the mean rank of the country with the lowest concentration of each variable to the mean rank of each other country. Error bars show standard deviation. For comparisons of the mean rank of each country with every other country, see Table S3. Number of “Bold” products per country; CA n = 1, JP n = 3, PL n = 2, ZA n = 1, UK n = 1; see Figure 1 legend for country abbreviations. All products were extracted and analyzed at least three times. * = statistically significant (p < 0.05) between the country with the lowest value and this country; see Results for exact p values. See Figure 2 legend for a key.