The scientific foundation for tobacco harm reduction, 2006-2011

A. Cancer

In 2007, the International Agency for Research on Cancer published Monograph 89 [2], which contained the findings from an IARC working group established in 2004. The committee found that there was sufficient evidence that ST use is carcinogenic to humans, with sufficient evidence that "ST causes cancers of the oral cavity and pancreas."

It is important to understand the magnitude of the risks codified by Monograph 89. Unfortunately, although the document contained detailed descriptions of some epidemiologic studies relevant to cancer among ST users, the monograph did not present any summary risk estimates. Instead, some individuals from the working group, led by IARC epidemiologist Paolo Boffetta, published a meta-analysis in 2008 concluding that ST use was associated with elevated risks for cancers of the oral cavity (Relative risk, RR = 1.8), esophagus (RR = 1.6) and pancreas (1.6), all of which were statistically significant [3] (Table 1).

Even though these risks are low in comparison with smoking, a 2009 meta-analysis published by Peter Lee and Jan Hamling did not confirm them [4]. In fact, after appropriate adjustment for confounding factors, the Lee-Hamling analysis found that ST use was associated only with prostate cancer (Tables 1 and 2). Lee and Hamling commented that "Prostate cancer is not considered smoking related [original citations removed], and more information on its relationship with ST is needed before any clear conclusion can be drawn."

Table 1 shows the major differences in risk estimates between the meta-analyses of Boffetta and that of Lee-Hamling for cancers of the oral cavity, esophagus and pancreas. A detailed subsequent analysis published by Lee and Hamling [5] revealed that the differences were due to the following factors: (a) Boffetta et al. did not use all available studies, and there were no specific criteria for which studies were included and which were excluded; Lee and Hamling used all available studies. (b) in some instances Boffetta et al. used only the highest risk estimates, even though they were derived from internally inconsistent subgroups of either ST exposure or of the disease outcome of interest; Lee and Hamling's analysis was conducted with clear and consistent criteria for ST exposure and disease outcomes [5].

The differences between the Boffetta et al. and Lee-Hamling meta-analyses are clearly illustrated in the case of pancreatic cancer, for which Boffetta reported a statistically significant summary RR of 1.6, and Lee-Hamling reported a summary RR of 1.07, which was not significant. The Lee-Hamling result was also reported in a more detailed analysis for pancreatic cancer published in 2008 by Sponsiello-Wang et al. [6].

Both meta-analyses used results from a 2005 Norwegian study [7] and a 2007 Swedish study [8] which reported risk estimates for pancreatic cancer among all snus users and for a subset of snus users who were never smokers. The Norwegian study reported a risk increase among all snus users (RR = 1.7, CI = 1.1 - 2.5) but not for the subset of snus users who were never smokers (RR = 0.9, CI = 0.2 - 3.1) [7]. The Swedish study reported exactly the opposite: There was virtually no risk among all snus users (RR = 0.9, CI = 0.7 - 1.2), but the subset of snus users who never smoked had an increased risk (RR = 2.0, CI = 1.2 - 3.3) [8].

As Lee pointed out in a recent review [9], "For pancreatic cancer, Boffetta cited only the increases for never smokers from the [Swedish] study and for the whole population from the [Norwegian] study, not mentioning the lack of increase for the whole population for the construction workers and for never smokers for the Norway cohorts." It is important to note that Boffetta was a co-author of both studies, which makes his selective use of data from them even more perplexing.

Ironically, the RR differences for pancreatic cancer between the Boffetta et al. and Lee-Hamling meta-analyses have been resolved in favor of Lee-Hamling by a third meta-analysis published in 2011, and it was co-authored by Boffetta [10]. It reported summary RRs for ever ST users (0.98, CI = 0.75 - 1.27), exclusive ST users (0.62, CI = 0.37 - 1.04), and ST users who smoked cigarettes (1.36, CI = 0.94 - 1.96). The authors concluded, "Our results on ST use are in broad agreement with a recently published meta-analysis of all published data on the issue, which reported no excess risk of pancreatic cancer in case-control studies" [6].

Other concerns have been raised about the validity of the risk estimates for pancreatic cancer from both the Norwegian [7] and Swedish [8] studies. The former was published in 2005 by Boffetta et al., and it was an extended follow-up of a cohort of Norwegian men recruited in the 1960s [7]. Boffetta et al. reported that ever users of ST had an increased risk for pancreatic cancer (RR = 1.67, CI = 1.12 - 2.50). However, despite the fact that alcohol consumption had been documented as the strongest risk factor for pancreatic cancer in this cohort, with odds ratios up to 10.8 [11], Boffetta et al. did not adjust the snus risk estimates for alcohol use [12]. In addition, Boffetta employed an unconventional classification for snus exposure and an unusual adjustment for smoking, as documented in a subsequent letter to the editor [12].

The Swedish study was published in 2007 by Luo et al., and it is one of many follow-up studies of Swedish construction workers published by investigators at the Karolinska Institute in Stockholm [8]. Luo et al. reported that current users of snus had an elevated risk for pancreatic cancer (RR = 2.1, CI = 1.2 - 3.6), but they excluded 135,000 construction workers recruited during the period 1971-74 because of "ambiguities" in questionnaire coding [13]. This cast considerable doubt about the credibility of a 1994 report by Bolinder et al., which stands almost alone in linking snus use with cardiovascular diseases [14]. The Bolinder cohort had contained only workers enrolled in the years excluded by Luo.

The Bolinder cohort [14] has been subjected repeatedly to inclusion and exclusion in ten studies published by Karolinska Institute investigators during the period 2005-2011 [8,15-23] (Table 3). The 2007 study by Hergens et al. [18], which found an elevated risk of fatal heart attack among snus users (RR = 1.28, CI = 1.06 - 1.55), is especially troublesome. This study excluded the Bolinder cohort, and Bolinder was a co-author, giving the impression that in 2007 Bolinder excluded her own cohort from 13 years earlier.

Hergens et al. [18] raised another serious and troubling question about Bolinder's characterization of snus use [14]. Bolinder defined "ST users" in her study as "subjects who reported only present ST use...," a clear representation of current use [14]. However, in describing the rationale for excluding the Bolinder cohort, Hergens et al. stated that "during the period 1971-1974 exposure information on snuff use was limited to ever or never use..." [18] The descriptions of snuff use in these two studies are conflicting and irreconcilable, and the only rational conclusion is that one is a misrepresentation. The credibility of the Karolinska Institute studies is contingent on the resolution of these discrepancies.

B. Cardiovascular Diseases

There have been at least ten epidemiologic studies evaluating the risks for cardiovascular diseases (primarily heart attack and stroke) among ST users. Two meta-analyses by Lee in 2007 [24] and Boffetta and Straif in 2009 [25] have provided summary RRs from these studies.

C. Other Diseases

D. Summary of Health Effects

1. Primary epidemiologic studies and subsequent meta-analyses do not provide convincing evidence that ST use is associated with cancers of the oral cavity, pancreas, and gastrointestinal tract. There is evidence from one meta-analysis that ST use is associated with a small risk for prostate cancer, although a biologic mechanism for this disease has not been established.

2. Epidemiologic studies have not documented that ST use is associated with significantly elevated risks for myocardial infarction; however, a small increased risk for stroke cannot be excluded.

3. Studies from Sweden document that ST use is not associated with benign gastrointestinal disorders and chronic inflammatory diseases.

4. Pregnant women who use snus are at risk for slightly smaller babies, and they also have modestly elevated risks for premature delivery, stillbirth and possibly preeclampsia. Although any form of nicotine should be avoided during pregnancy, the highest risks for the developing baby are associated with smoking.

A. Additional Evidence from Sweden

In 2006 Rodu and Godshall summarized the cumulative evidence from Sweden that ST, in the form of moist snuff called snus, has played an important role in low prevalence of smoking among Swedish men and, to a lesser extent, among Swedish women [1]. That year Ramström and Foulds reported the results from a population-based national Swedish survey from 2001-2002 [40]. They found that snus was the most common cessation aid among men, used by 24% during their last quit attempt. Among men who had used only one cessation aid, 58% had used snus, and the success rate among these subjects (66%) was significantly higher than that among those who used nicotine gum (OR = 2.2, CI = 1.3 - 3.7) or nicotine patch (OR = 4.2, CI = 2.1 - 8.6). Ramstrom and Foulds also found that "the odds of initiating daily smoking were significantly lower for men who had started using snus than for those who had not" (OR = 0.28, CI = 0.22 - 0.36). They concluded, "Use of snus in Sweden is associated with a reduced risk of becoming a daily smoker and an increased likelihood of stopping smoking."

In 2008 Furberg et al. investigated twelve variables and their interactions as correlates of smoking cessation among 14,700 regular smokers in the population-based Swedish Twin Registry [41]. They reported that ever use of snus was the strongest individual correlate for former versus current smoking (HR = 2.70, CI = 2.30 - 3.20). They concluded, "Swedes appear to be using snus as a form of nicotine replacement therapy despite a lack of clinical trials data to support its use as a smoking cessation aid."

B. Evidence from Norway

In 2008, the European Commission released a report entitled "Health Effects of Smokeless Tobacco Products." [44]. Except for one small part discussing THR (Section 3.8, pages 111-118), most of the report was very negative, even going so far as to deny that snus use has had any effect on smoking in Sweden and Norway.

The report concluded: "It is difficult to envision any significant impact of snus use on smoking cessation in Norway..." This was especially baffling, as Figures 19-22 (pages 42-43) show clearly that increased snus use over the last 20 years was concomitant with decreased smoking.

Norway occupies an interesting position in the European political arena, and in European tobacco issues. While it is located in the Scandinavian peninsula, shares a border with Sweden and has membership in the European Economic Area, Norway has twice rejected membership in the EU. Thus, it has not been subject to the EU ban on Swedish snus and similar smokeless products. In fact, information has emerged from Norway that the increasing use of snus in the past 20 years has resulted in a substantial decline in smoking among Norwegian men, a virtual reproduction of the Swedish experience reviewed by Rodu and Godshall [1].

Much of the information on THR in Norway has been produced by Dr. Karl Erik Lund, a respected tobacco researcher with SIRUS, the Norwegian Institute for Alcohol and Drug Research, an independent institution but also a government entity answerable to the Ministry of Health and Care Services.

In 2008, Dr. Lund gave a presentation on Norwegian tobacco use at the 51^st conference of the International Council on Alcohol and Addictions [personal communication]. He reported that among Norwegian men age 16-35 years, the prevalence of smoking declined from 50% in 1985, to 30% in 2007, while the prevalence of snus use increased from 10% to 30%.

Lund reported that snus is very popular as a quit-smoking aid among Norwegian men. Among those who quit smoking in 2007, snus was used by 23%, while nicotine gum was used by only 9%; the nicotine patch, Zyban and a quit line were used by even fewer (6%, 3% and 3% respectively).

Lund also presented information about the outcome of the last quit attempt by Norwegian male smokers age 20-50 years in 2007. Of those who used snus, 74% "quit smoking altogether" or experienced a "dramatic reduction in smoking intensity," which is very similar to the rate among American men in the 2000 National Health Interview Survey (NHIS) who switched to ST (discussed below in Section D). Lund added that the quit percentages for those who used nicotine gum, patch and Zyban were 50%, 47% and 40% respectively.

In 2009 Lund published a report on THR, in which he noted that existing anti-smoking measures will result in "diminishing marginal returns." [45] Lund provided a compelling rationale for THR, and noted in an epilogue that the Norwegian Health Directorate has cautiously endorsed the strategy.

In 2010 Lund et al. published a study confirming that Norwegian men prefer snus over all other methods to quit smoking [46]. The analysis was based on a survey by the Norwegian Institute for Alcohol and Drug Research, which asked 3,583 former or current smokers age 20-50 years what method they used when they last tried successfully (former) or unsuccessfully (current) to quit. Snus was used by 32% of all respondents, making it the most popular method by far. Other methods that enjoyed modest popularity were nicotine gum (18%), self-help material (12%), and the nicotine patch (10%). Nicotine inhaler, Zyban, Champix, telephone quit line, and help from health care professionals were also measured in the survey, but they had negligible usage rates.

Lund et al. reported an adjusted odds ratio (aOR) to indicate the effectiveness of ST products compared with nicotine gum, their reference product. For quitting completely, the aOR for snus was 2.7, meaning that it was nearly three times more effective than gum. Snus was also three times more effective than nicotine gum in "greatly reducing cigarette consumption" among continuing smokers (aOR = 3).

In 2011 Lund et al. also published a study comparing quit-smoking rates among snus users and never users in seven Norwegian surveys [47]. Quitting was defined as the percentage of ever smokers who were former smokers at the time of the survey.

As seen in Table 5, quit rates for snus users were always higher than for those who had never used snus; the results are statistically significant for all surveys except number 4. This provides compelling evidence that snus has played a powerful role in smoking cessation among Norwegians and, as Lund noted, it is consistent with the Swedish evidence.

C. Clinical Trials

Prior to 2006, only one clinical trial relating to THR had been conducted, a pilot study with one- and seven-year follow-ups [48,49]. But during the past five years, several more clinical trials have been completed.

In 2007 Mendoza-Baumgart et al. conducted small pilot trials focused on two ST products (either Exalt snus or Ariva dissolvable) versus a nicotine lozenge [50]. They evaluated toxicant exposure, subjective responses and product preferences among smokers using a cross-over design.

Mendoza-Baumgart et al. found that, compared with baseline smoking, all products produced significant reductions in 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), a metabolite of the TSNA 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), and all produced comparable effects on withdrawal and craving. Ariva was the preferred product, followed by the nicotine lozenge and Exalt. The authors concluded that "These findings make it difficult to ignore the potential of some ST products, specifically Ariva, to reduce exposure to [TSNAs], particularly NNK..."

In 2008 Sharp et al. reported the results of a nurse-led smoking cessation program, employing ad libitum use of alternative nicotine products including snus, among 50 patients undergoing aggressive treatment for advanced stage head and neck cancer [51]. As they noted, "Our aim was to support the patients to be smoke-free during and after cancer treatment, rather than to be nicotine-free."

Sharp et al. reported that 74% of patients were confirmed cigarette abstinent during the treatment period. Forty-one patients were alive at the one-year follow-up, and 28 were off cigarettes. Multiple product use was common; the nicotine patch was the most commonly used product (91%), followed by snus (54%).

In 2008 Tønnesen et al. published the results of an open, randomized smoking cessation trial using group therapy and Oliver Twist, a ST pellet made in Denmark [52]. Cessation at 7 weeks was higher among subjects using ST (31.5% vs. 19.2%, OR = 1.94, CI = 1.05 - 3.62), but no significant differences were seen at the 6-month follow-up. At six months, 17.5% of participants were still using ST, even though a trial objective was cessation at 12 weeks after enrolment.

The unimpressive response rate for ST may have been influenced by negative attitudes among study staff. Tønnesen et al. commented that "...we do not believe that the therapists induced a positive expectation in the ST group. Our impression is that the nurses were not convinced that ST would help or would be accepted by the smokers." Nevertheless, they concluded that the trial demonstrated short-term efficacy of ST in combination with group therapy, and they indicated that other trials were underway.

In 2010 Caldwell et al. compared snus with nicotine gum among heavy smokers in New Zealand [53]. After observing smoking patterns and consumption for one (lead-in) week, Caldwell's group gave 63 smokers three different cigarette substitutes, each for two weeks. The substitutes were Swedish snus (4-gram pouches in three flavors), Habitrol nicotine gum (4 milligrams of nicotine) and a peppermint pouch containing 4 milligrams of nicotine embedded in microcrystalline beads, produced by a Swedish company called Niconovum.

The researchers collected information from the participants about the "acceptability and the willingness of smokers to use" the substitutes. They asked five questions gauging satisfaction, and they reported that "subjects scored Zonnic and snus more highly than gum for four out of the five..." All three products significantly reduced craving for cigarettes, and all three "...enabled subjects to reduce their smoking significantly compared with the lead-in week."

Participants ranked Zonnic and snus higher than nicotine gum for both quitting and reducing smoking. "At the conclusion of the study, subjects were asked to rank the three products in order of overall preference. For their first choice, an equal number (40%) chose snus or Zonnic, while 20% chose gum."

In 2010 Carpenter and Gray published a small but powerful study documenting that dissolvable tobacco products "led to a significant reduction (40%) in cigarettes per day, no significant increases in total tobacco use, and significant increases in two measures of readiness to quit, either in the next month or within the next 6 months." [54]

The authors randomly assigned 31 smokers who were uninterested in quitting to receive Ariva or Stonewall dissolvable smokeless products, or to continue smoking cigarettes. Smokers were given "minimal instructions on how to use" these products and were "told that there is no safe tobacco product and that the best thing they can do for their health is to quit entirely."

They wrote that their findings suggest "that Ariva and Stonewall are effective products to curb withdrawal and craving," and that there is "no evidence that ST (Ariva or Stonewall) undermines quitting. To the contrary, readiness to quit (in the next month and within the next 6 months) significantly increased among smokers who used a ST product relative to those who continued to smoke conventional cigarettes." This addressed the concern that telling smokers about vastly safer smokeless substitutes will "undermine quitting."

In 2010 Cobb et al. compared the acute effects (within 45 minutes) of administration of Ariva, Marlboro snus, Camel snus, and Commit lozenges with own-brand and Quest (very low nicotine) cigarettes to smokers who had been abstinent overnight [55]. Each of the 28 participants attended seven sessions, during which 40 physiological and subjective measures were assessed at several time points after two separate administrations of each product.

Camel snus and Commit significantly decreased an "intention to smoke" measure after the second administration, although not as much as own-cigarette brand. Ariva and Marlboro snus did not affect this measure. Very similar findings were also observed for a craving measure.

These results may have been related to the plasma nicotine levels produced by the various products. Own-brand cigarettes resulted in sharp and significant increases in plasma nicotine, and Camel snus produced a significant increase 15 minutes after the second administration, but the other products did not.

Although Ariva and Marlboro snus had no effect, Camel snus and Commit exhibited modest subjective effects on abstinence symptoms, even though they were not as effective as own-brand cigarettes. Nevertheless, Cobb et al. made the sweeping conclusion that "currently marketed non-combustible [potential reduced exposure products] may not be a viable harm reduction strategy for US smokers."

In 2011 Kotlyar et al. published the results of a clinical trial which was done in 2006-7, assessing whether RJ Reynolds' Camel Snus and Philip Morris' Taboka (a precursor of Marlboro Snus) were viable substitutes for cigarettes [56]. They recruited smokers who were interested in quitting, assigning them to use one of three products: 4 milligram nicotine gum or lozenge, Camel Snus or Taboka (participants had a choice of various flavors for each product). Participants were instructed to use at least one or two doses of the assigned product per day during a one-week sampling period; for the next four weeks, they were told to use the product at least 6 to 8 times daily (and additional doses if needed). During week 5, participants were required to reduce consumption of the substitute; by the end of that week they had to be completely tobacco- and nicotine-free.

There were several interesting results. First, all participants in all groups had a reduction in exhaled carbon monoxide, clearly demonstrating that they smoked less than before the study. Participants in all groups had a reduction in the urine concentration of N'-nitrosonornicotine (NNN) and NNAL. The reductions were statistically significant except for NNN in Camel Snus users (p = 0.07).

Overall craving and withdrawal scores decreased over the 4 weeks in all groups, with no differences between the groups. Continuous abstinence rates over the 4 treatment weeks varied from 33% (Taboka) to 43% (Camel Snus). Two weeks after the treatment ended, 39% of the Taboka group, 47% of the Camel Snus group and 56% of the nicotine group were not smoking, but these percentages dropped to 23%, 31% and 33% respectively after ten weeks of complete abstinence. One possible reason for the precipitous drop in the smoke-free percentages was the insistence on abstinence after 4 weeks.

There were other interesting aspects of this study. The Taboka group smoked significantly more than those using nicotine or Camel Snus. It is possible that Taboka, which had very low nicotine levels, simply didn't satisfy smokers.

A total of 1,159 smokers responded to advertisements for the study. According to the researchers, 800 "were able to be reached and were screened over the telephone," and 429 qualified and were interested in participating. Another 212 did not show up for the orientation. The attrition didn't stop there: 211 smokers were enrolled in the study but only 130 were randomized to one of the three groups. Just 80 participants completed the 4-week treatment period and the one-week transition to abstinence.

These numbers represent one of the biggest challenges of clinical trials, especially in the field of risky behaviors like smoking. The 80 participants who completed the study represent only 6.9% of the smokers who originally responded, so they are an especially motivated group. That has been one of the problems with quit-smoking trials: It is impossible to generalize their results because the subjects are almost always derived from a highly selected population that is not representative of smokers in general [57].

In 2011 Barrett et al. reported the effects of "Swedish-style" snus and a 2 mg nicotine lozenge on delay of smoking and craving reduction among 15 smokers using four double-blind placebo-controlled sessions [58]. The investigators, who conducted the study at Dalhousie University in Halifax, Nova Scotia, reported that snus delayed the urge to smoke and suppressed craving only in the eight men in the study, but that snus was ranked as the least preferred product.

Although the investigators acknowledged the impact on smoking and craving, they believed that "the therapeutic potential of [Swedish-style snus] may be limited by its acceptability." However, there are some elements of their protocol that may explain participants' dislike of the products. First, subjects "were told the products may contain some of the ingredients commonly found in cigarettes (e.g., tar, ammonia, carbon monoxide, menthol, nicotine, sucrose, etc.)." It is unclear how the investigators communicated that non-combustible products contained "tar" and "carbon monoxide," but these statement likely affected product ratings. Second, although the subject of the paper was Swedish snus, the investigators picked a brand made in Denmark instead of much more popular and successful Swedish products. Finally, the preference ratings were probably affected by the fact that the snus was unflavored, whereas the nicotine lozenge was flavored with mint.

In summary, the study by Barrett et al. showed that snus "may have some therapeutic potential for those attempting to quit smoking," but it was designed in a manner that portrayed a very negative view of its acceptability.

In 2011 O'Connor et al. reported the results of a 2008 sampling study of Camel snus, Marlboro snus, Stonewall dissolvable pellets and Commit nicotine lozenges among 59 smokers not interested in quitting, 44 of whom completed the entire study [59]. After all products were sampled, 45% of participants chose Commit, followed by Marlboro (29%), Camel (14%) and Stonewall (12%), with 80% very or somewhat likely to purchase their preferred product in the next year. Over the seven-day trial phase, cigarette consumption decreased from 11.8 to 8.7 per day while use of the other products was constant at 4.7 units per day. This was accompanied by a decline in exhaled carbon monoxide.

Despite these findings, O'Connor et al. made some very negative conclusions: "However, we observed no true switching (i.e., abandoning cigarettes), even though [smokeless] and [nicotine replacement] products were provided without cost. It is clear that simply informing smokers of the lower risk and providing products is not going to result in major immediate shifts to smokeless alternatives. In the absence of some significant incentive, it is unlikely that information campaigns alone would lead to migration from use of cigarettes toward less hazardous nicotine sources among United States smokers."

It is unclear how O'Connor et al. generalized the results from 44 subjects in Buffalo, NY to all American smokers. Furthermore, although two sessions "presented information about the relative risks of ST and nicotine replacement products compared to cigarettes to provide a health rationale for considering these products as alternatives," the paper did not disclose the details about the extent and tone of the information, which may have played an important role in the results. Twenty-three of the subjects were women, who may have had a pre-existing bias against ST. In addition, 23 had already used pharmaceutical nicotine, which may partially explain the preference for Commit.

O'Connor et al. provided a valuable insight: "...the greater the range of products offered, the greater the proportion of smokers who may find a product they see as a viable substitute for cigarettes. This is consistent with the body of literature [original references omitted] suggesting that smokers' varied reactions to different products may be informative in themselves, meaning a 'sampling' approach may allow smokers to find an appealing alternative product to cigarettes."

D. American Survey Evidence

In 2008 Rodu and Phillips provided the first population-level evidence that American men have quit smoking by switching to ST [60]. Using data from the 2000 NHIS, which the CDC uses to estimate smoking prevalence in the U.S., Rodu and Phillips estimated that 359,000 American male smokers had tried to switch to ST during their most recent quit attempt, and 73% (261,000, termed switchers) were former smokers at the time of the survey, representing the highest proportion of successes among all methods. In comparison, the nicotine patch was used by an estimated 2.9 million men in their most recent quit attempt, but only 35% were former smokers at the time of the survey. Of the 964,000 men who had used nicotine gum, 34% became former smokers. Of the 98,000 men who used the nicotine inhaler, 28% quit successfully. None of the estimated 14,000 men who had tried the nicotine nasal spray became former smokers.

Rodu and Phillips showed that switching to ST compares very favorably with pharmaceutical nicotine as a quit-smoking aid among American men, despite the fact that few smokers know that the switch provides almost all of the health benefits of complete tobacco abstinence. The results of this study show that THR is a viable cessation option for American smokers.

In 2009 Biener and Bogen reported the results from the 2006-2007 Indiana Adult Tobacco Survey that had implications for THR [61]. Indianapolis had served as a test market of Swedish-style snus by the two largest American cigarette manufacturers. In 2006 Philip Morris launched a test market there for Taboka snus (in 2008 it was discontinued when Marlboro snus was launched). In 2007 Indianapolis was one of several expansion markets for RJ Reynolds' Camel Snus; that product went on to national distribution in 2009.

Biener and Bogen reported that almost 20% of survey respondents throughout Indiana were aware of snus. Awareness among smokers statewide was 44%, which was 4.5 times higher than awareness among non-smokers.

Awareness among respondents in central Indiana (i.e. around Indianapolis) was 29%. More importantly, about 64% of male smokers in central Indiana had heard about snus, and 20% had tried it. This is evidence that Philip Morris and Reynolds were targeting adult male smokers in their test-market campaigns, and that the manufacturers were fairly successful.

Biener and Bogen also reported that risk perception played an important role in getting people to try snus. Respondents who correctly believed that ST is less harmful than cigarettes were almost four times as likely to try snus as those who believed the misinformation about the differential risks. Unfortunately, this study revealed that 88% of all respondents incorrectly believed that ST was just as dangerous as cigarettes.

Biener and Bogen offer some perceptive comments on the sad state of smoker misinformation:

"Both marketing and health education messages should include the information that all tobacco products are harmful and that abstinence from all tobacco products is the most healthful choice. At the same time, simply saying that ST is 'not safe' is not a sufficient stance for public health communications. There is a recognized continuum of risk along which various tobacco products can be placed, with low-nitrosamine ST products much lower on the risk continuum than combustible tobacco, although it is not harmless. Devising an effective way to inform the public about the continuum should be an important research priority, as currently consumers are woefully incorrect in their assessments of relative risk of various tobacco products. This state of affairs could result in people deciding not to give up smoking in favor of a product lower on the risk continuum because they assume that all tobacco products are equally harmful."

E. Summary

There is extensive research evidence that ST use has been a key factor in the declining rates of smoking and of smoking-related diseases in Sweden. While it cannot be proven that the availability of ST would reduce smoking prevalence in other countries, the potential population health benefits of ST are far greater than the potential risks. This makes the continued prohibition of ST in cigarette-dominant markets an unsustainable and counterproductive public health debacle.

A. Evidence from Sweden

In 2006 Rodu and Godshall reviewed the published studies from Sweden documenting that there is no evidence that ST is a gateway to smoking, especially among youth [1]. This was confirmed in a 2008 study of 3,000 adolescents from the Stockholm area by Galanti et al. [62]. They found that "the majority of tobacco users (70%) started by smoking cigarettes" and "the proportion of adolescent smoking prevalence attributable to a potential induction effect of snus is likely small."

In 2008, the European Commission's Scientific Committee on Emerging and Newly Identified Health Risks concluded that "The Swedish data...do not support the hypothesis that...snus is a gateway to future smoking." [44]

B. Evidence from the U.S

Opponents of THR in the U.S. believe that it will lead to increased teenage ST use, which will function as a "gateway" to smoking [63]. It has been observed that teenagers who use ST are more likely than non-users to subsequently smoke [64-68]. But a close examination of the evidence suggests only that ST use is one of several behaviors associated with smoking, not that it leads to smoking.

In the U.S., concomitant use of cigarettes is common among ST users [69] (dual use will be covered in Section VI). However, investigators have not found credible evidence that ST use is a gateway to smoking among American youth. In 2003 Kozlowski et al. analyzed data from the 1987 NHIS survey and concluded that there was little evidence that ST use was a gateway to smoking, because the majority of ST users had never smoked or had smoked cigarettes prior to using ST [70].

The belief that ST is a gateway to smoking is based mainly on two longitudinal studies comparing subsequent smoking among adolescent ST users and non-users [68,71]. The first study, which used the 1989 Teenage Attitudes and Practices Survey and its 1993 follow-up, found that young males who used ST were significantly more likely to have become smokers at follow-up than non-users of tobacco (OR = 3.5, CI = 1.8 - 6.5) [68]. However, a subsequent analysis revealed that the earlier study did not take into account well-known psychosocial predictors of smoking initiation that were in the TAPS, including experimenting with smoking, below average school performance, household member smoking, depressive symptoms, fighting and motorcycle riding [72]. Inclusion of these variables into a multivariate model reduced the odds ratio of smoking among regular ST users to 1.7, which was not statistically significant. The investigators concluded that the earlier "analysis should not be used as reliable evidence that ST may be a starter product for cigarettes."

The second study found that 7^th and 9^th grade students who had used ST (in the past 30 days) were more likely than nonusers to be smoking two years later (OR = 2.6, 95% CI = 1.5 - 4.5), after controlling for smoking by family and friends, low grades, alcohol use and deviant behavior [71]. However, Timberlake et al. [73] have observed that regression analysis may not adequately control for imbalances in covariate distributions between ST users and nonusers. They analyzed data from the National Longitudinal Study of Adolescent Health after propensity score matching and found that adolescent ST use was not associated with an increased risk of smoking in later adolescence or young adulthood [73].

In 2005 O'Connor et al. examined data from the 2000 National Household Survey on Drug Abuse to determine if ST use led to smoking. They described the impact of ST use on subsequent cigarette smoking initiation as "minimal at best," and they concluded that the association of ST use and smoking seen in other reports "is likely a manifestation of dual experimentation rather than a causal relationship." [74]

In 2010 Rodu and Cole investigated the gateway issue by analyzing data from the 2003-05-07 National Survey on Drug Use and Health, which asked survey participants at what age they used cigarettes or smokeless for the first time [75]. Using this information, Rodu and Cole classified participants as cigarette initiators, ST initiators, or both, and they determined the prevalence of current smoking among these groups at the time of the survey. The analyses were restricted to white men age 18 years or over, and white boys aged 16 to 17 years, which are the groups most likely to have used ST.

The prevalence of current smoking among white men who were cigarette initiators was 35%. In comparison, the prevalence of smoking among ST initiators was only 28%, which was significantly lower (Prevalence ratio, PR = 0.80, CI = 0.77 - 0.84). The results for boys were even clearer. Current smoking among cigarette initiators was 43%, but only 18% of ST initiators smoked. This means that boys who had started with ST were less than half as likely to be smoking at the time of the survey (PR = 0.43, CI = 0.36 - 0.52).

Rodu and Cole concluded that "ST use has played virtually no role in smoking initiation among white men and boys, the demographic groups among which ST use is most prevalent. There is evidence that, compared with cigarette initiators, ST initiators are significantly less likely to smoke. This suggests that ST may play a protective role."

Despite the preponderance of the evidence, claims of a gateway effect persist, which prompted O'Connor et al. to note in 2005, "Continued evasion of the [harm reduction] issue based on claims that ST can cause smoking seems, to us, to be an unethical violation of the human right to honest, health-relevant information." [74]

C. Summary

It is now established that ST use is not a gateway to smoking in Sweden, nor in the U.S. In fact, there is evidence that the opposite is true: ST users may play a protective role against subsequent cigarette smoking.

A. Evidence

In 2010 that study was completed by Frost-Pineda et al, who reviewed 17 published research studies that had data on the health risks from dual use versus those from smoking [77]. Frost-Pineda and colleagues conclude that "...there are not any unique health risks associated with dual use of ST products and cigarettes, which are not anticipated or observed from cigarette smoking alone." The authors further commented that "some data indicate that the risks of dual use are lower than those of exclusive smoking."

Frost-Pineda et al. also reviewed longitudinal studies in the U.S. and Sweden to determine if dual users had a different trajectory of tobacco use and cessation than that of exclusive smokers. A 2002 study by Wetter et al. found that 11% of dual users were tobacco-abstinent after 4 years of follow-up, compared with 16% of exclusive smokers [69]. However, 80% of exclusive smokers were still smoking at the 4-year follow-up, while only 27% of dual users were smoking; 44% were still dual users and 17% were exclusive smokeless users.

Very similar results have been reported in longitudinal studies of dual users in Sweden. For example, Rodu et al. reported the follow-up tobacco status of men in northern Sweden who were either cigarette smokers or dual users when they enrolled in a population-based epidemiological study [78]. Among exclusive smokers followed for 5 years, 69% were still smoking, 4% were dual users, 7% used ST, and 19% were tobacco free; the respective percentages among dual users were 6%, 52%, 24%, and 18% [79]. Among smokers followed for 9 years, 51% were still smoking, 10% were dual users, 16% used ST, and 45% were tobacco free; the respective percentages among dual users were 4%, 44%, 41% and 11%. Among smokers followed for 13 years, 46% were still smoking, 7% were dual users, 12% used ST, and 36% were tobacco free; the respective percentages among dual users were 9%, 22%, 60%, and 9%.

Frost et al. concluded that, although dual users are less likely than exclusive smokers to be completely tobacco abstinent at follow-up, they are much less likely to be smoking.

B. Summary

In follow-up studies, dual users are less likely than exclusive smokers to achieve complete tobacco abstinence, but they are also less likely to be smoking. Follow-up studies also suggest that the health risks from dual use are lower than those from exclusive smoking.

A. Introduction

E-cigarettes are battery-powered devices that vaporize a mixture of water, propylene glycol, nicotine and flavorings. They are activated when the user inhales through the mouthpiece of the device. To date, most e-cigarettes and mixtures are manufactured in China.

E-cigarettes have been sold on the American market for several years. In 2008 and 2009, the U.S. Food and Drug Administration (FDA) detained shipments being imported by two American distributors, Smoking Everywhere and NJOY, on the grounds that the items were unapproved drug-delivery devices. The distributors filed a lawsuit in federal district court, and in January 2010 Judge Richard J. Leon ruled that the FDA does not have the authority to regulate e-cigarettes in that manner [80]. Judge Leon ordered that "the FDA shall not detain or refuse admission into the United States of [Smoking Everywhere's and NJOYs] electronic cigarette products on the ground that those products are unapproved drugs, devices, or drug-device combinations."

The judge found that the 2009 "Tobacco Act applies to 'tobacco products,' which Congress defined expansively as 'any product made or derived from tobacco that is intended for human consumption'...Congress enacted the Tobacco Act to confer FDA jurisdiction over any tobacco product - whether traditional or not - that is sold for customary recreational use, as opposed to therapeutic use. As such, the Tobacco Act, in effect, serves as an implicit acknowledgment by Congress that FDA's jurisdiction over drugs and devices does not, and never did, extend to tobacco products, like electronic cigarettes, that are marketed in customary fashion for purely recreational purposes."

Finding that e-cigarettes, like all tobacco products, are subject to FDA oversight but fall outside of both drug and device categorization, Judge Leon characterized the FDA's attempt to apply pharmaceutical standards to e-cigarettes as "bootstrapping run amuck."

The FDA filed an appeal, and in January 2011 an appellate court affirmed Judge Leon's decision requiring the FDA to regulate e-cigarettes as tobacco products, rather than as drug-delivery devices. On April 25, 2011 the FDA accepted the appellate ruling. In an open letter published on the agency's website, Tobacco Center Director Lawrence R. Deyton and Drug Center Director Janet Woodcock acknowledged that e-cigarettes are tobacco products and would be subject to regulations under the 2009 Tobacco Act [81].

The FDA decision was a victory on several counts for American smokers and for public health. First, the FDA decision guarantees that e-cigarettes, which have helped many smokers quit, will remain on the market. Second, the Deyton-Woodcock letter indicated that FDA regulation of e-cigarettes will subject them "to general controls, such as registration, product listing, ingredient listing, good manufacturing practice requirements, user fees for certain products, and the adulteration and misbranding provisions, as well as to the premarket review requirements for 'new tobacco products' and 'modified risk tobacco products.'" These requirements will promote the marketing of safe and quality-controlled products. Finally, the decision could allow pharmaceutical companies to reposition nicotine medicines as recreational alternatives to cigarettes. Today, these products are sold with a therapeutic claim for smoking cessation, but they are expensive, unsatisfying and FDA-approved only for temporary use (10-12 weeks). That accounts for their dismal success rate among smokers. Pharmaceutical companies may enter the recreational nicotine market with products that satisfy smokers indefinitely and are cheap enough to compete directly with cigarettes. Clearly, the tobacco industry is poised to compete in this market -- Reynolds American owns Niconovum (here http://www.niconovum.se/ ) and British American Tobacco recently formed Nicoventures (here http://www.nicoventures.co.uk/).

B. Scientific Studies of E-cigarettes

C. Summary

E-cigarettes produce a vapor composed of water, propylene glycol and nicotine, so e-cigarette users are not exposed to the thousands of toxic agents formed when tobacco is burned. Although laboratory studies have detected trace concentrations of some contaminants, this appears to be a small problem that is amenable to improvements in quality control and manufacturing that are likely with FDA regulation as tobacco products. There is limited evidence from clinical trials that e-cigarettes deliver only small doses of nicotine compared with conventional cigarettes. However, e-cigarette use emulates successfully the cigarette handling rituals and cues of cigarette smoking, which produces suppression of craving and withdrawal that is not entirely attributable to nicotine delivery.