Article Text

Download PDFPDF

Global surveillance of oral tobacco products: total nicotine, unionised nicotine and tobacco-specific N-nitrosamines
  1. Stephen B Stanfill1,
  2. Gregory N Connolly2,
  3. Liqin Zhang1,
  4. Lily T Jia3,
  5. Jack E Henningfield4,
  6. Patricia Richter5,
  7. Tameka S Lawler1,
  8. Olalekan A Ayo-Yusuf6,
  9. David L Ashley1,
  10. Clifford H Watson1
  1. 1Centers for Disease Control and Prevention, National Center for Environmental Health, Division of Laboratory Sciences, Emergency Response and Air Toxicants Branch, Atlanta, Georgia, USA
  2. 2Harvard University School of Public Health, Boston, Massachusetts, USA
  3. 3Centers for Disease Control and Prevention, National Center for Environmental Health, Division of Laboratory Sciences, Organic Analytical Toxicology Branch, Atlanta, Georgia, USA
  4. 4The Johns Hopkins University School of Medicine and Pinney Associates, Bethesda, Maryland, USA
  5. 5Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta, Georgia, USA
  6. 6Department of Community Dentistry, Faculty of Health Sciences, University of Pretoria, South Africa
  1. Correspondence to Stephen B Stanfill, National Center for Environmental Health, 4770 Buford Highway, Atlanta, Georgia 30341, USA; sstanfill{at}
  • Competing interests JEH serves as a consultant through Pinney Associates to GlaxoSmithKline Consumer Healthcare on an exclusive basis regarding matters relating to smoking cessation, has a financial interest in a potential new nicotine replacement product, and has provided expert testimony against the tobacco industry. The other authors declare they have no competing interests.


Objective Oral tobacco products contain nicotine and carcinogenic tobacco-specific N-nitrosamines (TSNAs) that can be absorbed through the oral mucosa. The aim of this study was to determine typical pH ranges and concentrations of total nicotine, unionised nicotine (the most readily absorbed form) and five TSNAs in selected oral tobacco products distributed globally.

Methods A total of 53 oral tobacco products from 5 World Health Organisation (WHO) regions were analysed for total nicotine and TSNAs, including 4-(methyl-nitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), using gas chromatography or liquid chromatography with mass spectrometric detection. Unionised nicotine concentrations were calculated using product pH and total nicotine concentrations. Fourier transform infrared spectroscopy was used to help categorise or characterise some products.

Results Total nicotine content varied from 0.16 to 34.1 mg/g product, whereas, the calculated unionised nicotine ranged from 0.05 to 31.0 mg/g product; a 620-fold range of variation. Products ranged from pH 5.2 to 10.1, which translates to 0.2% to 99.1% of nicotine being in the unionised form. Some products have very high pH and correspondingly high unionised nicotine (eg, gul powder, chimó, toombak) and/or high TSNA (eg, toombak, zarda, khaini) concentrations. The concentrations of TSNAs spanned five orders of magnitude with concentrations of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) ranging from 4.5 to 516 000 ng/g product.

Conclusions These data have important implications for risk assessment because they show that very different exposure risks may be posed through the use of these chemically diverse oral tobacco products. Because of the wide chemical variation, oral tobacco products should not be categorised together when considering the public health implications of their use.

  • Oral tobacco
  • nicotine
  • tobacco-specific N-nitrosamines
  • carcinogenicity
  • addiction potential
  • addiction
  • carcinogens
  • co-substance use
  • smokeless tobacco products
  • toxicology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Globally, oral tobacco products represent a diverse assortment of tobacco-containing products that deliver nicotine primarily through the oral mucosa upon placement in the mouth. These products may be chewed, sucked, or held between the gum and teeth for variable time intervals and, in some cases, swallowed in whole or part.1 2 Oral tobacco product use has varying geographic prevalence. An estimated 8.1 million people use oral tobacco products in the US; however, in Southeast Asia, an estimated 258 million people use oral tobacco products. In addition to its addictiveness, oral tobacco may contribute to diabetes, high blood pressure, cardiovascular disease, oral diseases, and cancers of the oral cavity and pancreas.1 3 Oral tobacco use is also associated with increased risk of death from myocardial infarction and increased risk of premature birth and pre-eclampsia.3 4

Oral tobacco products range from simple cured tobacco to elaborate products containing many non-tobacco ingredients; these products can be handmade or commercially made by using simple or very complex manufacturing processes.1 5 6 Some oral tobacco products contain significant amounts of plant material (betel leaf, areca nut, catechu, etc.); moreover, additives such as sweeteners, flavour agents and spices (saffron, cardamom, camphor, eucalyptus, etc.) are commonly added. Alkaline modifiers, including certain inorganic salts, slaked lime and ashes produced by burning certain wood (eg, Willow, Mamón) or fungi,1 5 6 are also added to some oral tobacco products. Unprocessed tobacco is mildly acidic (approx. pH 5–6.5); however, addition of alkaline modifiers boosts product pH converting a greater fraction of nicotine to more rapidly absorbed unionised nicotine,3 5 6 which contributes to faster spikes in blood nicotine concentrations7 and could result in such products being more addictive.7–9

Regional differences and local preferences contribute to the diversity of oral tobacco products in physical appearance, constituents and chemistry, with some products containing tobacco with little or no alkaline modifier, some augmented with substantial amounts of various alkaline modifiers,1 5 6 and some mixed with slaked lime (as the alkaline modifier) and areca nut. Some global oral tobacco products have unique characteristics, such as extremely high pH, nicotine-enriched tobacco (eg, Nicotiana rustica L.), non-tobacco plant ingredients (eg, catechu, betel leaf, spices, etc.) and alkaline modifiers (plant/fungi ashes; slaked lime, etc.) not associated with Western-style products (ie, snus, snuff, chewing tobacco).1 5 6 Moreover, some oral tobacco products, particularly those from South Asia (eg, betel quid, gutkha, mawa), also contain appreciable amounts of areca nut, seeds of the Areca palm (Areca catechu L.),5 6 which has been classified as an International Agency for Research on Cancer (IARC) group 1 carcinogen, although the actual carcinogenic constituent(s) has not yet been identified.10

Oral tobacco products are known to contain more than 30 carcinogens, including volatile aldehydes, lactones, polycyclic aromatic hydrocarbons (PAHs), heavy metals, radioactive metals and tobacco-specific N-nitrosamines (TSNAs).10 Among these chemical constituents in oral tobacco, TSNAs are considered the most potent classes of carcinogens11 with N-nitrosonornicotine (NNN) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), both IARC group 1 carcinogens,10 linked to the formation of cancers of the oral cavity, oesophagus, lung and pancreas.1 3 Another TSNA, 4-(methyl-nitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) has recently been reported in US moist snuff products.12 Clinical studies have shown that NNAL, a metabolite of NNK, is present in the urine of oral tobacco users at levels similar to those found in urine from smokers13; the NNAL present in some oral tobacco products12 may be absorbed directly and add to NNAL formed by metabolism of NNK.

Some products from South Asia, Sudan and South America contain N rustica, a tobacco species with higher nicotine and TSNA concentrations than found in cultivated tobacco used in US products (Nicotiana tabacum L.).14 In India, an estimated 35% to 40% of tobacco present in oral tobacco products is N rustica.15 16 Similarly, some Sudanese toombak contains very high concentrations of TSNAs17 likely from the use of N rustica. The type of tobacco and curing of oral tobacco products can influence nicotine content and carcinogenic TSNA content14 17 and, in turn, could potentially influence addiction and exposure to potent carcinogens.18

The primary goal of this study is to provide researchers and policy makers with approximate pH levels and total nicotine, unionised nicotine, and TSNA concentrations for a diverse global set of oral tobacco products. Because all products in this study were tested in a single laboratory using standard methodologies, direct comparisons between product groups are possible.


Sample overview

In total, 65 oral tobacco samples were purchased or obtained by research partners in the following countries of origin: Bangladesh, India, Nigeria, Pakistan, South Africa, Sudan, Sweden, Uzbekistan and Venezuela. In all, 12 samples that lacked detectable nicotine were excluded from the analysis list. This study explores the pH levels and the total nicotine, unionised nicotine and TSNA concentrations in 53 oral tobacco products from 5 WHO regions. The geographic distribution and common constituents of representative oral tobacco products from three broad categories, based on product constituents, is presented in table 1; photographs of selected products from each category are shown in figure 1.

Table 1

Description of representative products from three broad categories of oral tobacco products used globally1 3 5 6 (some products with the same name can fit in more than one category based on formulation)

Figure 1

Photographs of representative products from three broad categories of oral tobacco products used globally. Examples of category I are (A) pouch snus (Sweden), (B) tobacco leaf (Bangladesh), (C) natural leaf chewing tobacco (USA), (D) dry snuff (USA) and (E) low pH moist snuff (USA). Examples of category II are (F) chimó (Venezuela), (G) naswar (Uzbekistan), (H) khaini (India), (I) toombak (Sudan) and (J) medicated dry snuff (South Africa). Examples of category III are (K) handmade gutkha (India), (L) manufactured gutkha (India), (M) mawa (Pakistan), (N) mainpuri (Pakistan) and (O) zarda, areca nut-containing variety (Bangladesh).

In most cases, a list of locally popular types of oral tobacco products was identified in published reference documents15 and provided to the research partners. In some cases (eg, Uzbekistan), the research partner identified the type of oral tobacco product for testing. In all cases, the research partner was responsible for locating, purchasing and shipping the items to the Centers for Disease Control and Prevention (CDC) (Atlanta, Georgia, USA). Samples were not stored in the country of origin but were promptly shipped at ambient temperature to CDC; upon receipt, samples were labelled and stored in a freezer at −30°C until analysed to prevent product changes (ie, moisture loss), minimise loss of volatile constituents and inhibit the formation of TSNAs during storage. All pH, nicotine and TSNA measurements were performed in the Tobacco Analysis Laboratory at CDC.

Fourier transform infrared spectroscopy

These products differ widely in tobacco type, additives, non-tobacco constituents and production; in select cases, product composition and identification of tobacco (N tabacum and N rustica) and non-tobacco components (areca nut, alkaline agents) were aided by using Fourier transform infrared (FT/IR) spectroscopy. Samples were ground and homogenised in a handheld grinder prior to FT/IR analysis. Samples of N rustica, provided by the Great Lakes Inter-Tribal Epidemiology Center and the Wisconsin Native American Tobacco Network, and N tabacum, cultivated tobacco used in US products, were analysed by FT/IR spectra for comparison with international tobacco products. Chimó, a tar-like product from Venezuela, was applied to the attenuated total reflectance (ATR) crystal, analysed and subsequently removed with methanol. Analyses were performed by using a JASCO 6200 FT/IR spectrometer (JASCO, Inc.; Easton, Maryland, USA) fitted with a diamond crystal ATR (Pike Technologies; Madison, Wisconsin, USA). Absorbance mode spectral detection was accomplished by using a wide-band detector cooled to approximately –70°C with liquid nitrogen. Sample spectra were obtained by averaging 64 scans in the spectral range from 425–4000 cm−1 at 4 cm−1 resolution.

Total and unionised nicotine quantification

To quantify total nicotine, a 1 g sample of oral tobacco was extracted in 50 ml of methyl tert-butyl ether (containing quinoline as an internal standard) and 5 ml of 2 N NaOH. For analysis, 1 μl of the extract was injected into a gas chromatograph/mass spectrometer operated in selected ion monitoring mode19; analysis of nicotine was done in triplicate. Measurements of pH (±0.001 pH units) were performed by adding 2 g of oral tobacco product to 10 ml of distilled, deionised water. Many of the products analysed in this study were dry powders. In cases where the product absorbed all the water (resulting in a paste-like consistency), an additional 10 ml of deionised water was added prior to pH measurement.20 By substituting product pH and the pKa of nitrogen's pyrollic nitrogen (8.02) into the Henderson–Hässelbalch equation, the proportion of nicotine in the unionised form (αfb) was calculated. The amount of unionised nicotine was determined by multiplying total nicotine by the αfb value.20

Tobacco-specific N-nitrosamines quantification

The concentrations of five tobacco-specific N-nitrosamines ((1) NNN, (2) NNK, (3) N‘-nitrosoanatabine (NAT), (4) N‘-nitrosoanabasine (NAB) and (5) NNAL) in oral tobacco samples were measured in triplicate. Samples were ground, and a 0.25 g portion was transferred to an amber extraction vial and spiked with isotopically labelled internal standards. Samples were extracted with aqueous ammonium acetate on a Lab-Line shaker (Melrose Park, Illinois, USA) at 250 rpm for 1 h. Two quality control samples, Copenhagen moist snuff and 2S3 Reference tobacco (University of Kentucky; Lexington, Kentucky, USA), were prepared with each batch of samples. Extracts were filtered with a 0.45 μm nylon syringe filter and a 20 μl aliquot was injected on a Xterra C18 MS liquid chromatography column (Waters Corporation; Milford, Massachusetts, USA). Compound-specific detection was accomplished by using a triple quadrupole mass spectrometer operated under electrospray ionisation and multiple-reaction monitoring mode. All chromatographic data were processed by using Analyst 4.02 software from Applied Biosystems (Forest City, California, USA).


Product characterisation by FT/IR

Of the samples analysed, 25 had definitive product labelling. For example, several samples believed to be gutkha were analysed by FT/IR and the resulting interferograms were matched to known examples. Nut-like plant material from products presumed to contain areca nut were collected, washed and dried to remove other product constituents. The material was then examined by light microscopy. These samples were further analysed by FT/IR analysis and compared with spectra of fresh areca nut for confirmation. In this instance, the samples in question contained unique IR peaks corresponding to areca nut and were confirmed to be handmade gutkha.

Analyses of mawa and mainpuri by FT/IR confirmed the presence of areca nut. One zarda sample from Bangladesh contained areca nut while another zarda sample did not. Neither of the Indian khaini samples contained an FT/IR signature indicative of the presence of areca nut, a popular but non-essential additive in these products. Of the five Sudanese toombak samples received, one product was a coarse, tan-coloured powder that lacked detectable nicotine. Furthermore, FT/IR analysis revealed that this sample did not contain characteristic spectral peaks indicative of tobacco (N tabacum or N rustica); this non-tobacco product was excluded from further analysis. Another toombak sample, a dry, light brown product, contained nicotine and total TSNA levels three times higher than the other three toombak products, which were black and appeared similar to moist snuff. Determination of the type of tobacco present in the areca nut-containing zarda (Hakim Pury) was inconclusive due to interferences due to the areca nut present in that sample.

Because some oral tobacco products from Sudan contain N rustica, a nicotine-enriched tobacco variety,5 14 17 we compared the IR spectra of pure N rustica tobacco with each of the four toombak samples. A light brown toombak product (sample 5) exhibited spectral patterns corresponding to known samples of N rustica tobacco. The three other toombak samples had spectra similar to that found for products made with N tabacum, the species commonly used in US products. Some samples of gul and zarda with very high nicotine concentrations had spectral patterns most similar to N rustica.; however, these samples did not contain extremely high TSNA concentrations. Identity of tobacco species (N rustica or N tabacum) in tobacco samples was further confirmed by the ratio of IR-specific absorbances.

Product designations

Product designation categories used in table 2 are based on product constituents listed on the package labelling (if available), product pH and, in the case of some products, confirmation of the presence or absence of areca nut by FT/IR analysis and light microscopy.

Table 2

Levels of pH, nicotine and unionised nicotine found in international oral tobacco products


The pH in international oral tobacco products tested in the study ranged from pH 5.2 to 10.1, which translates to 0.2% to 99.1% of nicotine being in the unionised form (figure 2). The highest pH values were found in khaini (India), toombak (Sudan) and snuff (South Africa) (table 2). In terms of pH, handmade gutkha (pH 7.4–8.6) had a wider range of pH than the commercially manufactured gutkha (pH 8.5–8.9) analysed. The pH level in tobacco-only products (pH 5.2–7.2) was generally lower than oral products known to contain alkaline modifiers (pH 7.0–10.1) (figure 1), whereas areca nut-containing products ranged from pH 6.5–8.9.

Figure 2

The range of pH and percentage of unionused nicotine in various oral tobacco products from 10 countries from 5 WHO regions.

Total nicotine

Nicotine concentrations ranged from 0.2–34.1 mg nicotine/g product (mg/g) (table 2). Total nicotine in most products ranged from 0.2–21.3 mg/g; however, a few products, such as gul powder (Bangladesh), zarda (India), chimó (Venezuela) and toombak (Sudan), had higher nicotine concentrations ranging from 27.5–34.1 mg/g. One toombak sample (sample 5), with a FT-IR spectral pattern most similar to N rustica, had a nicotine concentration (28.2 mg/g) that was almost three times higher than the other three toombak samples (9.56–10.7 mg/g). Several other products, including Eagle Gul, Baba Zarda and tobacco leaf (Bangladesh), with higher nicotine values (19.7–33.4 mg/g) had FT/IR spectral features consistent with N rustica. Some chimó samples had high nicotine values (27.5–30.1 mg/g); however, FT/IR was inconclusive as to the tobacco type it contains.

Unionised nicotine

Unionised nicotine content, calculated by using product pH and measured total nicotine, spanned over four orders of magnitude (table 2). Calculated unionised nicotine concentrations for most products ranged from about <0.1–13.8 mg/g, except for two chimó products (27.4 and 30.1 mg/g) and two gul powder products (29.1 and 31.0 mg/g). Unionised nicotine was lowest in handmade gutkha from Pakistan (0.1 mg/g), Sada Pata tobacco leaf (0.2 mg/g) and wet zarda (0.2 mg/g). In terms of unionised nicotine concentrations, handmade gutkha (0.2–3.3 mg/g) was similar to the manufactured gutkha (1.0–1.8 mg/g) samples analysed in this study.

Tobacco-specific N-nitrosamines

The TSNA concentrations varied widely among the international samples (table 3). The highest concentrations of NNK were found in toombak from Sudan (516 000 ng/g). Dry zarda from Bangladesh had 3840 ng/g of NNK, much higher than most of the products tested. The highest concentrations of NNN were found in products from Sudan (368 000 ng/g), Bangladesh (28 600 ng/g) and India (18 600 ng/g) (table 3). Handmade gutkha and mawa from Pakistan contained the lowest NNK concentrations. Oral tobacco products contained a wide range of NNAL concentrations (3.58–6770 ng/g), unlike cigarette smoke, which does not usually contain detectable concentrations of this compound. The highest NNAL concentrations were found in four samples of toombak, and also in dry zarda and khaini.

Table 3

The concentrations of five tobacco-specific N-nitrosamines (TSNAs) found among various international oral tobacco products

All four nicotine-containing toombak samples from Sudan had high TSNA concentrations. This toombak product, with the highest NNK concentrations (516 000 ng/g) and extremely high nicotine (28.2 mg/g), was identified by FT/IR as containing N rustica. Zarda (Pakistan) and khaini (India) analysed in this study had very high TSNA concentrations. The NNN content in Zarda exceeded 28 000 ng/g and concentrations in khaini exceeded 17 000 ng/g. Among the gutkha products analysed, a handmade gutkha (product 1; Secunderabad, India) had the highest concentration of all five TSNAs, whereas a handmade gutkha from Pakistan had the lowest concentrations of the five TSNAs. The concentration of total TSNAs in the international products analysed in this study ranged from 83.9–992 000 ng/g (table 3).


Confirmation of product identity or composition using FT/IR was performed on 34 samples, including gul powder, tobacco leaf, zarda, khaini, gutkha, mawa, mainpuri, naswar, toombak, snuff, nasway, and chimó. Furthermore, FT/IR was used to determine whether a product contained tobacco similar to that used in U.S. products (N tabacum) or a higher nicotine-containing tobacco species (N rustica) or neither and whether or not it contained areca nut (A catechu). In a few cases, products contained a spectral pattern unlike either N tabacum or N rustica and may indicate the use of a different tobacco species (such as Nicotiana glauca Graham) in these products. Chimó is made by cooking tobacco, sodium bicarbonate, flavouring, brown sugar and Mamón tree ashes until the mixture becomes a concentrated black tar.1 5 Some products made in South America contain N rustica14; however, due to the tar-like consistence of chimó, FT/IR was inconclusive in determining the tobacco species present in these products. The high concentration of nicotine in chimó products is undoubtedly influenced by the nicotine content of the tobacco used and the preparation of the product.

The pH values in the international products (pH 5.2–10.1) (see figure 2) exceed the pH values found recently among top selling US moist snuff products (pH 5.5–8.6). Approximately 40% of the international products had pH values exceeding the highest value found for US moist snuff products (pH 8.6).12 Total nicotine among the international brands ranged from 0.16 to 34.1 mg/g product. For comparison, US moist snuff products range from 4.4–14.2 mg/g product with a single product as high as 25.0 mg/g product.12 Due to higher alkalinity and, in some cases, higher nicotine values, unionised nicotine had a much wider range (0.05–31.0 mg/g product) in many international products than found among US moist snuff products (<0.1–7.8 mg/g product).12

In this study, one toombak product had the highest concentrations of all five TSNA compounds, with NNN and NNK concentrations of 368 000 and 516 000 ng/g product, respectively. For comparison, the highest levels of NNN and NNK in US moist snuff products were 42 600 ng/g and 9950 ng/g product, respectively.12 The highest NNK levels in one toombak product was 50 times greater than the maximum concentration found among US moist snuff products.12 The total TSNAs (sum of all five TSNAs) in the international products analysed in this study ranged from 83.9–992 000 ng/g (table 3), whereas total TSNAs in US moist snuff ranged from approximately 4900–90 000 ng/g.12 A combination of factors, such as pH, tobacco type, nitrate fertilisation/uptake, curing, fermentation and storage conditions, could contribute to these extremely high TSNA concentrations. Moreover, salivary TSNA concentrations in the oral cavity of toombak users reach the low ppm (μg/ml) range,17 18 thus, it is not surprising that 68% of oral cancers in Sudanese men are attributed to the use of toombak or other oral products.3

In addition to toombak, several other products, including zarda (Pakistan) and khaini (India), also had very high TSNA concentrations compared with US moist snuff.12 Even though oral tobacco products, such as zarda, gutkha, or snuff, share the same name, in different countries the chemical composition can be different. Swedish snus had relatively low concentrations of most of the TSNAs, particularly NAB, NNK and NNAL. Although snus products purchased in South Africa had relatively low TSNA concentrations, Swedish snus products were at least four times lower. The concentrations of TSNAs in South African snus were also higher than a local South African snuff product (NTSU) not manufactured by using GothiaTek,21 the strict Swedish tobacco industry standards governing allowable toxicant content in snus. This observation, together with lower TSNA concentrations in other traditional products, suggests that such products could be produced with lower TSNA concentrations. Effective product regulation and testing and the dissemination of best manufacturing practices across nations, particularly with respect to manufacture of traditional products, could have a positive net effect in reducing carcinogen levels. The Surgeon General concluded that tobacco products should be no more harmful than necessary given available technology.22

Our findings confirm that TSNA levels vary widely in oral tobacco products. Factors, such as pH, nitrate content, tobacco type, curing, fermentation and storage conditions, which can be altered, could influence the TSNA content of a product. The data in this paper suggest that oral tobacco products can be produced with lower TSNA content. Efforts to reduce TSNAs in tobacco would likely reduce exposure to these known human carcinogens. Moreover, individual use factors, such as the type and amount of product used and duration of use, may affect the delivery, exposure and health risk. Oral tobacco products are highly diverse and present a complex array of potential health hazards. The use of cigarettes and oral tobacco makes estimating the cumulative exposure risks more difficult. Even among dedicated oral tobacco users, the availability and opportunity for using a wide variety of different oral tobacco product types makes risk assessment more challenging.

The concentration of unionised nicotine in an oral product is likely the primary characteristic that determines the extent of tobacco-dependent addiction that, in turn, results in repeated exposure to many harmful tobacco-related constituents during long-term use. The concentration of total nicotine alone may not adequately explain nicotine delivery and response. Alkaline agents can substantially increase nicotine absorption rates by converting nicotine to its most rapidly absorbed form. Moreover, use of nicotine-enriched tobacco (ie, N rustica) in some products (eg, gul powder, toombak) or processes that concentrate nicotine in a tar-like product (ie, chimó), could contribute to high nicotine concentrations, which in the presence of elevated pH, yield high unionised nicotine concentrations (figure 3). Potential modification of pH levels through addition of varying levels of alkaline modifiers could produce products with lower unionised nicotine levels suitable for initiation and products with progressively higher unionised nicotine and greater addiction potential that might facilitate product graduation.3 7 8 The wide concentration ranges seen among certain product types (eg, toombak, chimó, snuff) could help provide data useful on an empirical basis for specifying different or multiple maximum allowable concentrations for nicotine, pH and various toxicants as a function of product type by organisations such as WHO23 and the US Food and Drug Administration.24

Figure 3

Unionised nicotine concentrations (mg/g) in oral tobacco products from 10 countries from 5 WHO regions.


Oral tobacco products sent to CDC were a convenience sample available to our research partners at the time of the request and with their available financial resources; research partners were not reimbursed for the purchase or shipping of these products. These products do not represent an exhaustive sampling of individual countries geographically or a particular product group (eg, snuff, chimó, etc.). Due to wide variety of values found among the entire set of global products, analysis of one or a few samples of a particular product type gives a limited but informative view of product constituents. Further research with a larger set of products from each product group and more extensive sampling (ie, greater number of analyses) will be required to fully characterise these products so that they can be compared within a product group. In this study, sample measurements were made in triplicate. A greater number of measurements would be required to provide the level of statistical power necessary to make meaningful comparisons between individual products with very similar values. The results for these international products represent the analyte concentration in the products at the time of testing. The amount measured also does not translate directly into absorbed amount. The amount actually absorbed by users depends on numerous product characteristics, use parameters (e.g. amount used) and physiological differences in individual users. Even with these limitations, a clearer picture of oral tobacco as an inhomogeneous and diverse group of products has emerged from this global study.


Oral tobacco products are a chemically diverse group of products that can contribute to numerous health problems, including cancer and cardiovascular disease. Differences in the tobacco used, the various methods of curing and preparation, and the nature of other substances added to these products prior to use yield a varied group of products. When referring to such a diverse group of products, the term ‘oral tobacco’ is preferable to ‘smokeless tobacco’, as some of these products (eg, snuff, fire-cured dry snuff, iq'mik)1 5 6 are made using fire-cured tobacco that contains smoke-derived chemicals including phenols, PAHs and TSNAs also found in cigarette smoke.10 25

The global sample of oral tobacco products analysed here contained a wide range of pH levels and total nicotine, unionised nicotine, and TSNAs concentrations with some products containing NNN and NNK levels exceeding daily levels delivered in cigarette smoke.26 Tobacco products with higher unionised nicotine and TSNA levels generally leads to greater deliveries7 13 26 and, in some cases, may translate to higher risks for adverse health outcomes.3 Our data does not support oral tobacco products, as a class, being viewed as ‘safer’ or as providing a ‘reduced harm’ alternative to smoking. The possibility of dual use of tobacco products expands the potential for addiction and exposure to harmful constituents and may reduce the likelihood of complete abstinence from all tobacco products. At present, the only known means to reduce risk from tobacco is through cessation.

Oral tobacco products should not be routinely lumped together as a homogenous product category and considered as a single, equivalent product nor should their use be considered in isolation from other concurrent tobacco use. The drawing of broad conclusions about oral tobacco products based on limited data obtained on select samples from specific localities could be very misleading. Further studies to better characterise individual oral tobacco products, their diverse contents, the exposure of users to these products and the role of oral tobacco taken alone or in combination with other forms of tobacco are needed to provide crucial science to help inform consumers and also those involved in policy decisions and recommendations for tobacco control.

What this paper adds

  • Oral tobacco is often viewed as a homogenous set of products; however, on a global basis, oral tobacco products vary from simple tobacco-only products to those containing substantial amounts non-tobacco ingredients including various flavour additives or pH modifiers. Past studies have focused primarily on groups of related products or products from a particular geographical region or country. This study treats smokeless tobacco on a global basis to show the ranges of nicotine, tobacco-specific N-nitrosamines (TSNAs) and pH that exist among these heterogeneous products.

  • This paper is one of the first to examine nicotine, pH and toxic constituents across a diverse spectrum of oral tobacco products distributed globally. Because these diverse products were characterised in a single laboratory, the differences in concentration levels are meaningful and can be readily used to compare product categories. The observed wide concentration ranges of nicotine, unionised nicotine, TSNAs and pH values suggest that the impact on addictiveness, toxicity, or carcinogenicity of a given product type are not uniform and the oral tobacco products should not be lumped into a single category.


We would like to thank Dr. Ali Idris (Maritime Heart Center; Nova Scotia, Canada), Dr. Ricardo Granero (Ministry of Health and Social Development; State of Lara, Venezuela), Dr. K. Srinath Reddy (Public Health Foundation of India), Dr. Prakash Gupta (Helias; India), Dr. Nandini (India), Dr. Kiran Dhawan (India),Dr. Ayda Yurekli (WHO), Dr. Annette David (Health Partners; Guam), and other partners in Bangladesh, Nigeria, South Africa, and Sweden for obtaining and sending samples. We are grateful to the Great Lakes Inter-Tribal Epidemiology Center and the Wisconsin Native American Tobacco Network for providing us with Nicotiana rustica samples. We are also thankful to Madhu Chaudhary-Webb and Anand Ramaswamy for translating product labels from India, and to Mr. Phillip Ruer (JASCO, Inc.) for technical assistance with FT/IR. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.



  • Funding This work was funded by the U.S. Government, Department of Health and Human Services. This study was also funded internally at the Centers for Disease Control and Prevention, with funds directly provided by the U.S. federal government.

  • Provenance and peer review Not commissioned; externally peer reviewed.