Annex L
Scientific Committee on Tobacco and Health Technical Advisory Group
Review of Emissions in Cigarette Smoke
Introduction
1. For more than 20 years successive Governments have taken various courses of action to encourage smokers to stop smoking and non-smokers not to start. At the same time a programme of product modification has allowed smokers, unable to give up, to smoke products with reduced emissions. This paper summarises a review undertaken by the Technical Advisory Group (TAG) of the Scientific Committee on Tobacco and Health (SCOTH) to examine the impact on health of this programme, updating earlier work.1,2,3
Tar
2. The product modification programme, and subsequent EU regulations, have had a significant impact on tar yields over the past few decades. Typical UK cigarettes had tar yields of about 25-35 mg per cigarette during the 1950s and about 5-15 mg per cigarette in 1990.4 The sales weighted average tar yield of cigarettes has shown a continuous steady decline through the last three decades, from 20.8 mg in 1972 to 11 mg in 1993.5 UK Regulations, based on an EC Directive, placed an upper limit of 15 mg of tar per cigarette from the end of 1992, falling to 12 mg from the end of 1997.
3. Although lower tar cigarettes still cause health problems, there is reasonably good evidence to show that, with respect to lung cancer, lower tar cigarettes are less carcinogenic than higher tar cigarettes. For example, the decrease in male lung cancer mortality in England and Wales since the 1960s is consistent with an effect of reduced tar yield, beyond a simple reduction in male smoking prevalence. In 1988 the Independent Scientific Committee on Smoking and Health (ISCSH) concluded that "Smoking lower tar cigarettes confers a reduced risk of lung cancer than does the smoking of cigarettes with the relatively high yields that were customary twenty five or more years ago".1
4. Two studies which examined the impact on health of reduced tar have provided further evidence. Tang et al, combining mortality data from a number of prospective studies in the UK in relation to the tar yield of cigarettes, showed that about one quarter of deaths from coronary heart disease, lung cancer, and possibly other smoking related diseases, could be avoided by switching from historically high 30 mg tar cigarettes to those yielding 15 mg tar.6 Parish et al assessed the impact of smoking cigarettes of different tar yields on the incidence of non-fatal myocardial infarction (MI) and showed that the overall relative risk of suffering a non-fatal MI is 10% higher for smokers of medium tar cigarettes (mean tar yield 13.3 mg per cigarette) compared with smokers of low tar cigarettes (mean tar yield 7.5 mg per cigarette).7
5. Both studies investigated smokers who were likely to have smoked products with higher tar yields earlier in their lives. The impact of tar reduction could, therefore, be greater when comparing risks in lifetime smokers of higher and lower tar cigarettes. However, it cannot be stated strongly enough that tar reduction is no substitute for not smoking, as the adverse impact of smoking on disease regardless of cigarette type is clearly demonstrated in each of these studies.
6. When considering the effectiveness of the tar reduction programme, information on dosimetry should be taken into account. Exposure to tobacco smoke may be estimated by measuring levels of cotinine, a nicotine metabolite, in blood, saliva or urine. Some compensatory smoking of lower-yielding products does occur, but unless smokers deliberately remove filters or block ventilation holes, there is increasing difficulty in compensating up to former doses of nicotine and consequently of tar.8
7. The prevalence of smoking is strongly linked with socio-economic status, being highest in more deprived groups.9 There are indications that such factors also influence brand choice and smoke uptake. Socio-economic factors exert independent effects on morbidity and mortality and this could affect the interpretation of data on relationships between yields of tar and relative risks for smoking related diseases.
8. Over the last few years there has been a steady increase in the proportion of smokers reporting they smoke low tar brands of manufactured cigarettes. In 1986 some 19% smoked low tar brands (under 10 mg per cigarette) increasing to 25% in 1992 and 32% in 1994.9 This trend is due partly to deliberate switching by smokers and partly to the decline in tar yields of all manufactured brands.
9. The increase in popularity of low tar brands is partly offset by an increased prevalence in the smoking of hand-rolled cigarettes. Between 1986 and 1992 prevalence was steady, with 18% of male and 2% of female smokers smoking mainly hand-rolled cigarettes. In 1994 this increased to 21% and 4%, respectively.9 Recent studies show that a significant proportion of tar yields from hand-rolled cigarettes can be greater than the current maximum limit of 15 mg per cigarette for commercial cigarettes,10,11 although due to the variable nature of the product, measurement of tar yields from hand-rolled cigarettes is not straightforward. A study undertaken by the Laboratory of the Government Chemist10 showed that few UK hand-rolled cigarette smokers use filters and high-porosity papers, both of which have played an important part in reducing tar yields in commercial products.
Nicotine
10. Nicotine is addictive and it is this fact which is largely responsible for the continuation of smoking and the consequent exposure to the harmful effects of tar. The perpetuation of the smoking habit results in a wide range of health effects. No previous or current UK regulations have addressed nicotine yields directly. Under current Labelling Regulations, the yield of nicotine, as well as tar, is specified on packets, but no upper limit is set for nicotine. In mainstream smoke nicotine is present in the particulate phase and most cigarette design measures, used to control tar, should also control nicotine. However, for much of the 1970s and 1980s, whilst average tar yields were falling, average nicotine yields did not fall in proportion.1 This appears to have been brought about through blending tobaccos, increasing the proportion of material with a high nicotine content.
11. During the 1990s machine measured nicotine yields fell more sharply than in previous years. Revised analytical methods introduced in 1990 account for part of this trend, but it is probable that control of tar yield meant it was no longer possible to maintain nicotine yield at previous levels.5
12. The ISCSH expressed the view that, apart from its important addictive properties, nicotine, in the doses delivered by smoking, is not thought to be harmful to a healthy individual, although it may have adverse effects on people with cardiovascular disease12. However, nicotine is a precursor of some of the tobacco-specific N-nitroso compounds, although the role of these substances in the development of smoking-associated cancers is unclear.13Nicotine may also be implicated in the deficit in birth-weight associated with smoking in pregnancy.
13. The main cause of concern about nicotine is its addictive properties. Nicotine dependent smokers seek an optimum nicotine dose and if yields are reduced are liable to compensate by smoking more cigarettes or by smoking each cigarette more intensively thus negating some of the advantages of tar reduction.8
14. Apart from compensatory smoking nicotine uptake is also affected by smoke quality; notably pH. The smoke from most UK manufactured cigarettes is acid, leading to deep inhalation for more effective nicotine absorption. Nicotine in the alkaline smoke associated with some continental and American brands is absorbed more effectively in the mouth and upper respiratory tract. However, deep inhalation is still required to produce the rapid pulse of nicotine to the brain associated with absorption in the lung alveoli.
15. When considering the benefit of nicotine reduction, two conflicting issues emerge. Firstly, in nicotine dependent smokers, compensatory smoking following nicotine reduction might negate some of the health advantages of further tar reductions. Secondly, as nicotine is accepted as the prime addictive agent in cigarettes, nicotine levels should fall at least as fast as tar to reduce the likelihood of addiction. This point is important for young smokers, who, although likely to continue to experiment with smoking, would be less likely to become addicted, and suffer adverse long term health effects, if nicotine yields were very low.
16. These two issues are difficult to reconcile. The yield of nicotine, as measured by a smoking machine, is not the main determinant of nicotine addiction. The bioavailability of inhaled nicotine, depth of inhalation, and the kinetics of absorption, distribution in the body, metabolism and excretion of nicotine are all important and cannot be determined or compared by a smoking machine. Experimental studies also indicate that the rapid peaks in blood nicotine levels, experienced on exposure, are important in the pharmacology and addictiveness of nicotine14, and will not be eliminated simply by a reduction in machine-measured nicotine yields.
17. It seems logical, though conjectural, that non-addictive levels of nicotine in cigarettes exist, but there are currently limited data to show what such levels might be. The work of Benowitz15 suggests a non-addictive level of uptake of 0.1-0.2 mg of nicotine per cigarette which would be the maximum bioavailability from cigarettes with a total content of 0.5 mg of nicotine. Some current UK low tar brands have smoking machine yields of 0.5 mg of nicotine per cigarette and below but the total nicotine content of these brands is several milligrams and thus significantly greater than the 0.5 mg suggested by Benowitz. Whether people would smoke cigarettes with "non-addictive" levels of nicotine has not been tested and they are likely to have low consumer acceptability.
18. It is important to consider the possible impact of reductions in nicotine yield and changes in the ratio of tar:nicotine yields on existing smokers. Large numbers of smokers are addicted to nicotine and altering nicotine levels in middle range cigarettes will not change this. Lowering nicotine yields may increase the health risks of existing smokers who compensate by smoking more, thereby increasing their intake of tar. On balance, however, there is no clear indication regarding advantages or disadvantages of varying the tar:nicotine ratio and the preferred option may be to ensure that tar and nicotine yields fall in line with one another.
19. Issues for consideration if control of nicotine yields is contemplated are:
- the role of nicotine in smoking-related diseases,
- the role of nicotine in addiction and establishing an addictive threshold,
- factors influencing smoke uptake, such as pH, smoking patterns and nicotine absorption.
The Health Survey for England could provide a valuable means of monitoring what actually happens in terms of nicotine uptake in the population at large. In some of these annual surveys cotinine determinations have been made on the blood samples taken from participants. So far, data are available from the surveys in 1993 and 1994,16 and show no appreciable change in cotinine levels among smokers over a one year interval. Continuation of the series could demonstrate to what extent compensatory smoking offsets the potential benefits of the further reductions anticipated to meet the new tar limits from 1998. Further insight into yield/uptake relationships would be obtained if brands smoked were also recorded for all subjects providing blood samples.
Carbon Monoxide
20. Carbon monoxide is a poisonous gas and contributes nothing to the flavour or "satisfaction" of smoking. It is however an inevitable component of the slow combustion process of tobacco smoking. Although CO has not been regulated, yields of CO have declined, but at a slower rate than tar yields. Some methods of tar reduction, notably filtration, have little or no effect on yields of CO and other gases. Other methods that dilute smoke, such as increased paper porosity and use of ventilation holes, reduce gases as well as particulates. However, yields of CO have fallen less than those of tar over the past few decades, tending to increase the CO:tar ratio.1,5
21. Smoking is the commonest reason for raised carboxyhaemoglobin (COHb) levels in the blood and direct or indirect measurements of COHb are widely used as a test for recent (same day) smoking.17 Levels in smokers are typically around 6-7% but can rise to 11% or more. Acute cerebral effects occur at much higher levels, (e.g. malfunctioning heating appliances or exposure to car exhaust in confined spaces) and are usually the result of accidents. Levels due to smoking have been shown to be within the range seen in experimental studies to reduce the time to angina pain in exercising patients suffering from heart disease.18 Endogenously produced carbon monoxide within the central nervous system and in the circulation may serve as a regulator of normal function. Adverse effects of carbon monoxide may, therefore, be linked to the high levels of exposure seen with accidental poisoning but there remains uncertainty of its role at low levels of chronic exposure.
22. Any further action to reduce tar is likely to involve measures such as increased ventilation or changes in the porosity of the cigarette paper, both of which will have a diluting effect on CO. Consequently, separate action to reduce CO yields is probably unnecessary.
Oxides of nitrogen.
23. In addition to carbon monoxide, nitric oxide (NO) is a notable constituent of the gas phase of tobacco smoke. Nitric oxide is produced by the decomposition of nitrates in tobacco, rather than by the combination of atmospheric nitrogen and oxygen, as occurs in sources such as gas flames or internal combustion engines. In cigarettes the oxide produced initially, and inhaled by the smoker, is NO, whereas other sources of oxides of nitrogen usually produce some nitrogen dioxide (NO2). Although NO exhaled by the smoker or contained in sidestream smoke, once dispersed in a room, will gradually oxidise to NO2, the smoker receives only NO. The physiological effects of NO and NO2 are quite different. Inhaled NO appears to have no direct toxic effect whilst NO2 is a respiratory irritant. However, recent research suggests that smoking adversely affects the physiological function of endogenous NO.19
24. NO in cigarettes is related to tobacco type rather than cigarette design. For a given tobacco blend, NO is reduced, along with CO and other gases, by increasing ventilation or paper porosity. Blends containing air-cured tobacco, which is relatively rich in nitrates, produce higher yields of NO than flue-cured tobaccos.1
25. The relationship between yields of carbon- and nitrogen-derived noxa and tar is interesting. Work conducted for the former ISCSH showed that yields of carbon-derived noxa followed those of tar, reasonably closely, correlation coefficients being between 0.6 and 0.9 in the survey of 75 brands.1,20 This relationship should now be re-examined as tar yields have reduced by around 20% since the work was undertaken. Yields of nitrogen-derived noxa are independent of tar yields and relate to type of tobacco and its nitrate content.
26. Apart from nicotine, carbon monoxide and oxides of nitrogen, there are over 4,000 compounds present in the gaseous and particulate phases of tobacco. The specific role of other noxa in smoking-related diseases is poorly understood and research to link yields of other noxa to smoking-related diseases is limited. In addition, experimental toxicity data on other compounds present in tobacco smoke is often based on routes of administration other than inhalation, making extrapolation of results difficult.
27. In the 1960s the American tobacco industry embarked on a programme to reduce the carcinogenic effects of cigarette smoke by reducing levels of carbon-derived polynuclear aromatic hydrocarbons (PAHs) and related compounds in tar. This was achieved not only through cigarette design parameters, but also by using blends of tobacco rich in nitrate which inhibit the formation of these compounds during pyrolysis. Experimental evidence shows that the tar from cigarettes made from nitrate-rich tobacco and containing reduced yields of PAHs and phenolic compounds is less carcinogenic than that from tobaccos rich in carbon and low in nitrate.21 While data from US epidemiological studies show smaller increments in lung cancer mortality in relation to dose, e.g. the number of cigarettes smoked per day,22 it is not clear whether this is linked to differences in the characteristics of the tobacco or to differences in smoking behaviour, like discarding longer butts.
28. Air-cured tobacco blends are less popular with UK smokers, but consideration should be given to whether there could be any potential benefits in moving towards such blends. Nitrate-rich tobaccos yield higher levels of N-nitroso compounds than other tobaccos. Although tobacco specific nitrosamines are potent carcinogens in animals, their role in the development of smoking-associated cancers in humans is unknown.
29. A large amount of data on the effects of tobacco cultivation, cigarette design and manufacture was taken into account during the course of this review and additional research has been carried out since publication of these data. Modern processing technologies, such as homogenised leaf curing, expanded tobacco and reconstituted tobacco sheet production, may provide scope for both quantitative and qualitative reductions in tobacco smoke emissions. The introduction of synthetic tobacco substitutes into cigarettes, which was the subject of extensive toxicological work in the late 1960s and early 1970s,1 proved to be a commercial failure, despite promising results in terms of reduced carcinogenicity in animal experiments, but other radical changes in cigarette design might be contemplated by industry.
Conclusions
30. Having considered the data available on emissions in cigarette smoke, the Technical Advisory Group came to the following conclusions.
- Tar:- The tar content of cigarette smoke is the single most important factor in terms of health risk. Tar reduction reduces the risk of developing some smoking-related diseases, notably lung cancer. However, this is small and does not compare with the benefit derived from giving up smoking altogether. Consumer acceptability of lower tar cigarettes is increasing, suggesting scope for further tar reduction, although the smoking of high yielding roll your own products is increasing and should be monitored.
- Nicotine:- The principal harmful attribute of nicotine is its addictiveness, which maintains the smoking habit, and thresholds of addiction have been postulated. Modest reductions in nicotine yields are unlikely to prevent addiction. It is considered extremely important, therefore, to continue to discourage young people from experimenting with cigarettes in order to prevent addiction occurring. Although limiting nicotine to non-addictive levels to prevent new smokers from becoming addicted is appealing it is unlikely to be practical. Most existing smokers are addicted to nicotine and drastic reductions in nicotine yield may have little impact on smokers' health risks, because of compensatory smoking. Because the link between tar and smoking related diseases is clear, further reductions in tar, with a proportional decrease in nicotine, are recommended.
- Further consideration of the optimum ratio of tar:nicotine is required. More information on the bioavailability, pharmacology and pharmacokinetics of nicotine and factors that influence them, such as smoke pH, is needed in order to attract attention to the possible benefits.
- A programme to monitor actual uptake of nicotine in relation to changing machine yields is called for, and this could conveniently be based on cotinine measurements within the Health Survey for England.
- Carbon monoxide:- The carbon monoxide yield from cigarettes has decreased over the past 25 years, but at a slower rate than that for tar. As further measures to reduce tar yields are likely to result in similar reductions in CO, no specific action is necessary. However the ratio of yields of CO to tar should be kept under review.
- Other noxa
- Research suggests that NO in tobacco smoke has a detrimental effect on the physiological function of endogenous NO. Further developments in this area will be watched with interest. As with CO, tar reduction measures are likely to result in reductions in NO.
- Yields of carbon-derived noxa, such as PAHs and aldehydes, closely follow those of tar. Yields of nitrogen-derived noxa do not, however, and are related to the nitrate content of tobacco. These relationships should be examined in representative current brands.
- The role of specific other noxa in the pathogenesis of smoking-related diseases is generally unclear. However, the carcinogenic role of PAHs and the inverse relationship of carbon- and nitrate-derived noxa requires further investigation, though conducting animal studies in this field is now difficult.
- Tobacco Industry: Dialogue with the Tobacco Industry will help clarify recent developments in tobacco cultivation and processing and cigarette design and their relationship to the production, modification and control of emissions from tobacco smoke.
|
