Nicotine products relative risk assessment: a systematic review and meta-analysis

Systematic literature review

Systematic searches were conducted using specific search terms pertaining to the health risks of nicotine products on July 7^th, 2020 in the MEDLINE (Pubmed) and NIH clinical trials (ClinicalTrials.gov) databases (see extended data¹¹). Due to the broad scope of the searches, the most relevant literature was targeted by searching at the title and abstract levels. The publication lists returned by the searches were exported and screened at the title, abstract and full-text levels according to pre-defined inclusion and exclusion criteria (see extended data¹¹). The screening steps were completed by one researcher and confirmed by a second. During each screening step, the reason for exclusion of each individual publication was recorded and the level of evidence assessed for the final shortlisted publications using an adapted version of the Oxford evidence-based medicine level of evidence scale (see extended data¹¹).

Data extraction and harmonization

The shortlisted studies were analyzed in detail to extract health risk data and relevant meta-data, including the nicotine product, brand, disease/symptom, methodology used, measurement and unit, error in measurement, significance of measurement, geographic location, sample size and conflict of interest. The extracted data were subsequently grouped by study and data type into epidemiological, toxin emissions, biomarkers of exposure, biomarkers of effect and adverse events groups. Within these groupings, the studies were clustered by nicotine product into the following 13 categories: combustible cigarettes, cut tobacco, chewing tobacco, dipping tobacco, snus, cigars, cigarillos, western pipe tobacco, water pipe tobacco, non-tobacco pouches, heat-not-burn devices, electronic cigarettes and nicotine replacement therapy.

Data analysis

The extracted and harmonized dataset was analyzed in three main segments: lifetime cancer risk, biomarkers of exposure and epidemiological data. The latter two segments are further subdivided into cancer and non-cancer risk.

Lifetime cancer risk. The lifetime cancer risk (LCR) of each nicotine product was calculated separately for inhalable and smokeless products. The inhalable products LCR calculation was based on the methodology outlined by Stephens⁹, whereas the LCR of the smokeless products was determined using the methodology outlined by the FDA¹⁰.

The LCR of each inhalable nicotine product was calculated from the toxin emissions data by adjusting the OEHHA unit risk values¹². From the full toxin emissions dataset, the International Agency for Research on Cancer (IARC) Group 1 carcinogens (“known carcinogens”) were selected, making a total of 13 carcinogens that were included in the analysis¹³. For each carcinogen, the OEHHA unit risk values were sourced and converted from risk per μg per m³ to risk per μg per breath by assuming the average breath volume of a healthy human is 500 mL¹⁴. The toxin emissions of the nicotine products were reported in varying units. For instance, combustible cigarette studies reported toxin emissions as μg or ng per stick, whereas electronic cigarette studies reported toxin emissions as μg or ng per 150 puffs. Therefore, the toxin emissions data for each product were converted to per puff values. In order to make this conversion, the average number of puffs per product/session was extracted from puff topography studies in the scientific literature (see extended data¹¹). The cancer potency of each nicotine product was calculated by adjusting the unit risk values with the observed masses of toxins in the emissions from each inhaled nicotine product using Equation 1:

Equation 1: Cancer potency of the nicotine product

Where P_i is the cancer potency of the ith nicotine product, C_i,j is the mass of the jth toxin in the ith nicotine product and U_j is the unit risk for the jth toxin. The cancer potency (P_i) represents the excess cancer risk associated with continuous lifetime use of each nicotine product. To put the cancer potency values into real-world context, the lifetime cancer risk was calculated by adjusting the cancer potency values for average consumption patterns of each product, using Equation 2:

Equation 2: Lifetime cancer risk of inhalable nicotine products

Where LCR_i is the lifetime cancer risk of the ith nicotine product, P_i is the cancer potency of the ith nicotine product, D_i is the average daily number of puffs taken by users of each nicotine product and B is the average number of breath taken in one day (40,000 breaths, equivalent of 20 m³ breathed per day). The lifetime cancer risk (LCR_i) represents the excess cancer risk associated with average daily use of each nicotine product over the course of a person’s lifetime.

For the non-inhaled (smokeless) products, the estimated lifetime cancer risk (ELCR) equation as defined by the FDA was used¹⁰. This equation calculates the lifetime cancer risk based on adjustment of the cancer slope factor for each carcinogen with the observed amounts of toxins measured in smokeless tobacco products and average consumption of the products:

Equation 3: Estimated lifetime cancer risk of smokeless nicotine products

Where ELCR_i is the estimated lifetime cancer risk of the ith product, C_i,j is the concentration of the jth toxin in the ith product, IR_i,j is the intake rate of the jth toxin in the ith product, AB_i,j is the absorption rate of the jth toxin in the ith product, EF_i,j is the exposure frequency of the jth toxin in the ith product, ED_i,j is the exposure duration of the jth toxin in the ith product, BW is the body weight of the average user, AT is the averaging time of use and CSF_j is the cancer slope factor of the jth toxin.

Biomarkers of exposure analysis. The excess biomarker levels of six IARC classified Group 1 carcinogens and 14 U.S Environmental Protection Agency (EPA) classified non-cancer toxin biomarkers in the users of each nicotine product, relative to non-users, were determined.

All values included in the analysis had been corrected for urine creatinine concentration, a common reference biomarker for urinalysis, and represented users of only one, single nicotine product. The average levels of these biomarkers in a population of non-nicotine product users were used to derive an excess biomarker level for each product using Equation 4. Where there were gaps in the data, the non-smoker referent value of 1 was set as a default. This assumes that nicotine products would never actively reduce toxin biomarkers in their users, therefore, the minimum (background) level of toxin biomarker will always be equivalent to that found in non-users. The excess biomarker level (EBL) represents the excess amount by which the users of each nicotine product are exposed to systemic toxins above the levels found in a non-user:

Equation 4: Excess biomarker levels in nicotine product users

Where EBL_i is the excess biomarker level for the ith nicotine product, P_i,j is the biomarker level in nicotine product users for the jth toxin biomarker and R_j is the reference level for the jth toxin biomarker.

Epidemiological data analysis. Risk ratios, odds ratios and hazard ratios were extracted from the epidemiological studies and a set of meta-analyses were performed to determine the relative risk of cardiovascular disease, respiratory disease, cancer and mortality in users of each nicotine product compared to non-users of any nicotine products. The epidemiological data extracted from the systematic literature searches was screened to include only relative risk values that compared current users of a single nicotine product to non-users of any nicotine product. Relative risk values were excluded if they were unadjusted for tobacco smoking. Where more than one study was available for a specific disease, the best available evidence was selected according to the level of evidence scale (see extended data¹¹). For instance, if a meta-analysis of prospective cohort studies and a single case-control study were available, the former would be included and the latter would be excluded. The lowest level of evidence included in the analysis was case-control studies; cross-sectional studies were excluded.

The remaining data after screening were grouped by disease type into cardiovascular disease, respiratory disease, cancer and mortality. The cancer category was further broken out into oral, other head and neck, lung, gastrointestinal and other. Meta-analyses of the relative risk data for each disease category were conducted using a random-effects model in the Comprehensive Meta-analysis (CMA) software Version 3.3, with a statistical significance threshold of α = 0.05. The meta-analysis could also be conducted in the open-access alternative, the metaphor package of R. A final meta-analysis was conducted to obtain overall relative risk ratios of cancer and non-cancer diseases for each nicotine product.

Relative risk hierarchy

The RRH combines the results of the lifetime cancer risk and epidemiological analysis, with a weighting system that accounts for the level of evidence and completeness of the dataset. In order to integrate these two analyses into a combined risk score for each nicotine product, an arbitrary scale from 0 to 100 was defined, with 0 representing non-users of nicotine products and 100 representing users of combustible cigarettes. Combustible cigarettes were selected as the top of the scale because they were the highest risk product in both of the analyses. Non-user groups were the control or baseline in both analyses, therefore no correction was required to the lower end of the scale. After converting both analyses onto a 100-point scale, a weighting of 5 was applied to the lifetime cancer risk analysis and 3 to the epidemiological analysis. The weighting scale is based on the 5-point scale that was used to classify the level of evidence, with systematic review/methodological syntheses of the data representing the top of the scale, followed by prospective cohort studies, retrospective cohorts studies, case-control studies and cross-sectional studies (see extended data¹¹). Two points were deducted from the epidemiological analysis in the weighting due to the fact that data were not available for all of the nicotine products and the quality of evidence was more variable, with single cohort and case-control studies being the best available for most of the data points. Alternatively, toxin emissions data was available for every nicotine product and the unit risk/cancer slope factors are based on large, comprehensive systematic literature review studies completed by OEHHA.

Sensitivity analysis

The sensitivity of the RRH to each analysis was determined by simulating several weightings of the lifetime cancer risk and epidemiological data and assessing their outcomes on the risk hierarchy. The analyses were weighted 5:3, 1:1 and 3:5, producing three simulations of the RRH.

Statistical software

Data was extracted from the scientific literature into a Microsoft Excel Version 16.41 spreadsheet. Microsoft Excel was used to conduct the lifetime cancer risk calculations. Comprehensive Meta-analysis (CMA) software Version 3.3 was used to conduct statistical meta-analyses in the epidemiological data-analysis. Calculation of the combined risk scores that were incorporated into the RRH was completed in Microsoft Excel.

Literature review

Systematic searches of the academic literature for studies investigating the health risks and hazards associated with nicotine products returned a list of 3,980 publications (Figure 1). Of these, 2,824 were excluded at the title screening step, 209 were excluded at the abstract screening step and 665 were excluded at the full text screening step. This left 320 publications that were included in the analyses, of which 53 were included in the quantitative analyses and RRH. During the screening steps, additional exclusion criteria were defined and added to those originally outlined in the methodology. The most common reasons for exclusion across all the review stages were “efficacy for smoking cessation only”, “not related to nicotine products”, “related to marketing or sales of nicotine products”, “related to prevalence of usage of nicotine products” and “related to consumer perceptions of nicotine products.” A full list of additional exclusion criteria can be found in the extended data¹¹.

Figure 1. Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) flow diagram.

NP – nicotine product.

Data analysis

The toxin emissions data was available for the full spectrum of nicotine products and the unit risk and cancer slope factor values provided a direct, quantitative link to cancer risk. Epidemiological data was available for over half of the nicotine products, and also provided a direct measure of disease risk. Biomarkers of exposure data were available for the majority of nicotine products, but the link between the concentration of the biomarker and disease risk was less clear compared with the toxin emissions and epidemiological data so this analysis was excluded from the RRH. Similarly, the biomarkers of effect and adverse event reports were excluded from further analysis due to the heterogeneity and subjective nature of the datasets, respectively.

Lifetime cancer risk analysis

The lifetime cancer risk was calculated using toxin emissions or content data for the inhalable and ingestible nicotine products, respectively (Table 1). The toxin emissions/content data was extracted from 12 publications as ranges of masses, which vary depending on the sub-type, brand or variety of nicotine product. The combustible cigarettes and heat-not-burn device datasets were the most complete, with 100% of the values available in the literature. For cut tobacco, cigarillos, cigars and water pipe tobacco, 40 – 60% of the data points were available, and any gaps in the data were filled with combustible cigarette values, based on the assumption that all combustible tobacco products would have a comparable emissions profile. For the electronic cigarette and nicotine inhaler, 50% and 75% of data points were not available, respectively, and no assumptions were made to fill these gaps due to the lack of comparable products. Data were available for all carcinogens in each smokeless product.

Cigarillos and water pipe tobacco have the highest cancer potency with values of 0.95 ± 0.013 and 0.94 ± 0.04 respectively, compared with 0.93 ± 0.004 for combustible cigarettes and cigars (Table 2). The adjustment for consumption causes significant differentiation between the combustible products. Combustible cigarettes and cut tobacco occupy the top the hierarchy for lifetime cancer risk, with 3,490 ± 16 and 3,464 ± 12 excess cancer cases per 100,000, and 650 ± 3.04 and 645 ± 2.33 times the risk compared to the nicotine inhaler, respectively (Figure 2). Cigarillos, which have a lower daily consumption, are associated with 2,938 ± 41 excess cancer cases per 100,000 and 547 ± 7.6 times the risk compared with the nicotine inhaler. Cigars and water pipe tobacco are also consumed less regularly and carry risks of 1,767 ± 26 and 1,748 ± 70 excess cancer cases per 100,000, and 329 ± 4.92 and 325 ± 13 times the risk of the nicotine inhaler, respectively.

Figure 2. The lifetime cancer risk of 13 nicotine products relative to nicotine replacement therapy (NRT).

The lifetime cancer risk of each nicotine product is shown here relative to NRT. The error bars on the chart represent the ranges of toxin emissions for different variants of each product. The data completeness (DC) is marked as a percentage in the x-axis.

After the combustible products, the ingestible products have the next highest cancer potency. Dipping tobacco and chewing tobacco have cancer potency values of 0.27 ± 0.17 and 0.12 ± 0.02, compared to 0.09 ± 0.02 and 0.08 for snus and non-tobacco pouches, respectively. When adjusted for consumption, heat-not-burn devices place higher in the hierarchy than the ingestible products, with 118 excess cancer cases per 100,000, compared to 25 ± 16 for dipping tobacco and 11 ± 2.3 for chewing tobacco. Similarly, non-tobacco pouches score lower than electronic cigarettes after adjusting for consumption, with 7.8 excess cancer cases per 100,000, compared with 8.2 ± 6.2. Relative to the nicotine inhaler, dipping tobacco, chewing tobacco, snus and non-tobacco pouches carry 4.6 ± 3, 2.1 ± 0.42, 1.6 ± 0.28 and 1.5 times the lifetime cancer risk, respectively.

Among the next generation inhalable nicotine products, heat-not-burn devices have the highest cancer potency (0.02), followed by electronic cigarettes (0.002) and the nicotine inhaler (0.0004). When adjusted for consumption, heat-not-burn devices are associated with 118 excess cancer cases per 100,000, compared with 8.2 ± 16 for electronic cigarettes and 5.4 for the nicotine inhaler. This equates to 22 times the risk of a nicotine inhaler for heat-not-burn devices and 1.53 ± 0.37 times the risk for electronic cigarettes.

Epidemiological analysis

Overall, 101 risk ratios across eight of the nicotine products were included in the analysis (Table 3). No studies that met the inclusion criteria were identified for cigarillos, non-tobacco pouches, electronic cigarettes, heat-not-burn devices or nicotine replacement therapy (Table 3). The completeness of the data for most products was above 80%, with only chewing, dipping and water pipe tobacco below. It should be noted that a large number of epidemiological studies were identified that investigated smokeless tobacco products, however, relatively few studies differentiated between chewing and dipping tobacco and no assumptions were made to fill the gaps in the data here.

The risk ratios relative to non-users for each nicotine product, obtained by meta-analysis of the cancer and non-cancer disease risk ratios extracted from the scientific literature, are listed in Table 4. In all cases, the risk ratio for cancer is greater than the risk ratio for non-cancer risk, but the confidence intervals are much larger (Figure 3). The risk ratios vary from 1.1 (1.04 – 1.17) for snus to 3.3 (2.44 – 4.43) for combustible cigarettes. For cut tobacco, to compensate for the lack of studies that specifically reported on cut tobacco, the combustible cigarette values were assumed on the basis that the majority of studies did not differentiate between factory-made and roll-your-own cigarettes in their classification of “cigarette smokers” and cut tobacco is primarily consumed in the form of roll-your-own or make-your-own cigarettes.

Figure 3. The risk ratios of 8 nicotine products in the epidemiological analysis.

The risk ratios are plotted here on a bar chart with the dark grey bars representing cancer risk and the light grey bars representing non-cancer risk. The data completeness for each product is shown as a percentage in the x-axis. The color of the bars indicates the completeness of the data, with darker colors representing more complete and lighter colors representing less complete datasets. The error bars represent the 95% confidence intervals.

In the cancer group, the highest risk products are combustible cigarettes and cut tobacco (RR = 3.28, 2.43 – 4.43), followed by water pipe tobacco (RR = 2.64, 1.64 – 4.26), western pipe tobacco (RR = 2.38, 1.28 – 4.41), dipping tobacco (RR = 2.06, 1.38 – 3.09), chewing tobacco (RR = 1.81, 1.04 – 3.17), cigars (RR = 1.68, 1.21 – 2.32) and snus (RR = 1.12, 0.89 – 1.4). All of the cancer risks are statistically significant (p < 0.05), with the exception of snus (p = 0.292), suggesting no statistically significant difference in the cancer risks between snus users and non-users of nicotine/tobacco products based on the available data. The non-cancer group follows the same overall order, with lower risk ratios compared to the cancer group and the absence of water pipe tobacco. Combustible cigarettes and cut tobacco have risk ratios of 1.97 (1.6 – 2.4), followed by western pipe tobacco with 1.86 (1.65 – 2.13), dipping tobacco with 1.31 (1.15 – 1.48), chewing tobacco with 1.22 (1.14 – 1.31), cigars with 1.20 (1.08 – 1.34) and snus with 1.11 (1.04 – 1.17). Here, all of the risk ratios are statistically significant (p < 0.05).

Relative risk hierarchy

The combined risk scores derived from the lifetime cancer risk and epidemiological analyses are plotted on a bar chart to form the RRH (Figure 4). In the RRH, the combustible products occupy the top end of the spectrum, with scores of 100 (91.3 – 111.2), 99.5 (90.9 – 110.6), 84.2 (83 – 85.3), 75.7 (56.7 – 107.3), 55 (47.1 – 66.7) and 41 (38.6 – 44.3) for combustible cigarettes, cut tobacco, cigarillos, western pipe tobacco, water pipe tobacco and cigars, respectively. Combustible cigarettes and cut tobacco are almost equivalent, followed by 10 – 20% decreases in risk for each of the proceeding products; cigarillos, western pipe tobacco, water pipe tobacco and cigars. Dipping and chewing tobacco follow the combustible products with combined risk scores of 15.1 (11.2 – 20.6) and 11.2 (8.01 – 16.5) respectively, representing a 63% drop in risk compared with cigars. Heat-not-burn devices and snus carry 3.3 and 3.5 times less risk than chewing tobacco, respectively. Finally, at the lower end of the spectrum, electronic cigarettes, non-tobacco pouches and the nicotine inhaler score less than 0.25 and only marginally elevated compared with non-nicotine product users, which have a score of 0 on the scale.

Figure 4. The relative risk hierarchy of the 13 nicotine products.

The combined risk scores for the 13 nicotine products are represented on the hierarchy. The combined risk scores were determined by combining the lifetime cancer risk and epidemiological analyses. The error bars represent a combination of the range of nicotine product emissions from the lifetime cancer risk analysis and the 95% confidence intervals for the epidemiological data. The data completeness (DC) is marked as a percentage in the x-axis.

The sensitivity analyses (see extended data¹¹) showed that the overall order of the RRH is robust to different weightings of the analyses. The order of snus and heat-not-burn devices inverse when greater weight is given to the epidemiological analyses, and this is equally true for electronic cigarettes and non-tobacco pouches. The most sensitive products in the risk hierarchy are snus, chewing tobacco, dipping tobacco, cigars and water pipe tobacco, with deviations in their combined risk scores of 15 – 60%.

Combustible tobacco products

The combustible products are the highest risk products across both analyses and in the final RRH. The combined risk score for cut tobacco is 99.5 (90.9 – 110.6), making it almost identical to combustible cigarettes, the high-risk referent at 100 (91.3 – 111.2). In the toxin emissions analysis, only 20% of the values for cut tobacco were available from studies that directly measured toxin emissions from this product. The remaining 80% of data points were filled with the combustible cigarettes values, based on the assumption that cut tobacco is generally consumed as ‘roll-your-own’ or ‘make-your-own’ cigarettes, and that the emissions would be comparable. Indeed, the few data points that were available from cut tobacco studies supported this assumption, showing very similar toxin emissions between roll-your-own and factory-made combustible cigarettes^16,23,24. Similarly, in the epidemiological analysis, the combustible cigarette risk ratios were assumed for 100% of the cut tobacco data points, based on the fact that most epidemiological studies did not differentiate between factory made and roll-your-own cigarettes and defined only “cigarette smoker” or “non-smoker”, thus encompassing both products^27,31,35,40. In addition, the consumption patterns and toxin emissions for these products are very similar and so it is reasonable to assume that the health outcomes would also be comparable.

In the toxin emissions analysis, data points that were missing for the combustible products were filled with the values reported for combustible cigarettes. This was based on the assumption that the toxin emission profile of combustible cigarettes would be roughly comparable to all products based on combustible tobacco. This assumption was based on the fact that where toxin emissions data was available for combustible tobacco products, they were comparable to or greater than the values for combustible cigarettes (Table 1)^16–19. Therefore, combustible cigarette emissions represent a reasonable minimum for combustible tobacco emissions. The data that was available for combustible products supported this assumption. The key differentiator between the combustible products is the adjustment for consumption patterns, which determines the order of the combustible products. According to the toxin emissions data, cigarillos and water pipe tobacco have a higher cancer potency than combustible cigarettes, however, cigarillos are consumed at a rate of 5.4 per day and water pipe tobacco had an average consumption of 3 sessions per week, whereas combustible cigarettes are consumed at a rate of 15 per day. On the other hand, daily users of cigars consume 4 per day on average, which decreases the overall lifetime cancer risk making this product the lowest risk among combustibles. This is in agreement with the epidemiological data, which also shows a reduced risk of cigars compared with other combustible products (Table 2)^{29,33,37,39,50,67}. The assumptions regarding consumption patterns are based on consumer survey data of daily users of each product provided by Euromonitor International.

Dipping and chewing tobacco

Dipping and chewing tobacco carry less risk compared with the combustible products. This is generally consistent across the two analyses, with the exception of cigars in the epidemiological analysis, which place slightly lower than dipping and chewing tobacco. Conversely, in the toxin emissions analysis there is a large difference between all of the combustible products and the smokeless tobacco products. This difference holds in the RRH, with dipping and chewing tobacco placing significantly lower than the combustible products, but remaining discernably elevated compared with the reduced risk products.

Reduced risk nicotine products

The reduced risk category is comprised of heat-not-burn devices, snus, electronic cigarettes, non-tobacco pouches and nicotine replacement therapy, all of which score lower than 5 on the 0 to 100 scale of the RRH. This category can be further divided into the tobacco-based reduced risk products and the tobacco-free reduced risk products. The former consists of heat-not-burn devices and snus, which have combined risk scores of between 3 and 3.5, and the latter consists of electronic cigarettes, non-tobacco pouches and NRT, which have combined risk scores below 0.25.

With the exception of snus, the reduced risk products are not represented at all in the epidemiological analysis, which can be attributed to their relative novelty compared with combustible and smokeless tobacco. In the toxin emissions analysis, the heat-not-burn devices place higher than chewing and dipping tobacco, however, this order is reversed in the final hierarchy due to the position of chewing and dipping tobacco relative to combustible cigarettes in the epidemiological analysis. Nicotine replacement therapy and electronic cigarettes lack 75% and 50% of data points in the toxin emissions analysis, respectively. Therefore, these products could potentially shift in the hierarchy as more data comes to light.

There are two key limitations in the placement of the next generation nicotine products in the hierarchy. The first is the lack of comprehensive and high-quality data, which leaves significant gaps in the analyses. The second is the rapid evolution of these products. Electronic cigarettes are currently on their fourth generation of development since their introduction to the United States market in 2007, having undergone changes to the atomizer unit, battery and other components, which can affect the aerosol⁸¹. It is highly important, therefore, to clearly identify the exact brand and generation of the device used in future research; this could also permit further breakdown of the categories for next generation devices and focus their development on reduction of risk to the user.

Biomarkers of exposure analysis

The biomarkers of exposure analysis (see extended data¹¹) was not included in the RRH due to the lack of a direct link between the biomarker levels and health outcomes, and the high inherent variability of this data. The high standard errors associated with biomarkers of exposure data have been characterized elsewhere in the scientific literature and include differences in external exposure level, chemical characteristics of the biomarker and differences in the elimination half-life of the chemical⁸². In the biomarkers of exposure analysis, all combustible products as well as chewing and dipping tobacco are within the margins of error of one another. Equally, snus and nicotine replacement therapy are both within the margin of error of non-users of any nicotine products. Overall, the large errors associated with the biomarker levels of each product user make it impossible to confidently order the products on a relative risk assessment scale.

Consistency with other studies

As outlined in the introduction, there are three main studies in the scientific literature that have ranked nicotine products in terms of their relative risk using different methods^7–9. The relative harm spectrum of nicotine products developed by Nutt and colleagues in 2014, and later adapted by Abrams and colleagues in 2018, places combustible cigarettes at the top of the spectrum^7,8. Following combustible cigarettes, small cigars, pipes, cigars and water pipe tobacco were listed. This order of combustible products is identical to the order presented in this study, with the exception of cigars and water pipe tobacco which are inversed. The actual score of small cigars, or cigarillos, is also very similar. However, after cigarillos, the Nutt and Abrams scales score the remaining combustible products between 10 and 25, which represents a major difference in comparison to this analysis. Here, the remaining combustible products score between 40 and 75. The smokeless tobacco products and reduced risk products, on the other hand, have comparable scores and placement, although the breakdown of the products is slightly different with Nutt’s classification differentiating refined and unrefined smokeless tobacco and various forms of nicotine replacement therapy.

Stephens created a risk spectrum based on the cancer potencies and lifetime cancer risk of tobacco smoke, heat-not-burn devices, electronic cigarettes and the nicotine inhaler⁹. The methodology used in the lifetime cancer risk analysis presented in this study is based on the methodology used by Stephens and the overall order of the products is consistent. Furthermore, the cancer potency ratios for each product shows excellent agreement with the values determined by Stephens.

In summary, this study is largely consistent with previous studies that characterize the relative risk of nicotine products published in the scientific literature, with only a few minor alterations to the overall order and significantly higher risk associated with all combustible products, compared with smokeless and reduced risk products. Nevertheless, this work represents an advancement in the relative risk assessment of nicotine products due to the systematic and data-driven methodology, applied to a broad spectrum of nicotine products and focused on the best available evidence in the scientific literature.

Limitations of the study

The key limitation of this study lies in the availability of the input data. In producing the RRH, we have highlighted where there are key gaps in the literature, particularly for the next generation nicotine products.

Another limitation of this analysis is the static view of consumption. While it was imperative to account for consumption patterns in order to make the risk hierarchy meaningful in the context of real-world product usage, this necessitated selecting one specific consumption level for each product and incorporating it into the analyses. As a result, the risk of certain products may be under- or overestimated in cases of lower or higher than average consumption. For example, the consumption of combustible cigarettes is assumed to be 15 per day in this analysis, which is associated with a specific level of risk. A smoker of 40 cigarettes per day may be at significantly higher risk than represented in this study, whereas a smoker of 1 cigarette per week may be at lower risk. Furthermore, this analysis only accounts for consumption of a single nicotine product; dual and poly-product usage risk data was excluded. Therefore, an individual consuming multiple nicotine products at the same time, may be at a different level of risk than that associated with consumption of any one product.

The combined relative risk values in this study represent an average across an ensemble of individual diseases. In doing so, we generalize disease risk to cancer and non-cancer risk, and then to a combined risk score. The risk values are only meaningful at this level and cannot be applied to individual diseases. This means that certain specific indications may be under- or over- represented by the relative risk values. For instance, the combined risk ratio for cancer, calculated from the epidemiological data, is 3.283 for combustible cigarettes, whereas the individual risk ratio for lung cancer is 8.96 and 1.3 for stomach cancer. Therefore, the combined risk value is meaningless at the level of individual cancers and can only be applied to the full ensemble. This is also true for the non-cancer combined risk scores.

Underlying data

All data underlying the results are available as part of the article and no additional source data are required.

Extended data

Open Science Framework: Nicotine Products Relative Risk Assessment: A Systematic Review and Meta-analysis. https://doi.org/10.17605/OSF.IO/3GX42¹¹

This project contains the following underlying data:

Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).