Revue d’Épidémiologie et de Santé Publique 56 (2008) 286–290
Translation of a
(Translated October 11, 2008 by Iro Cyr and Pr. Robert Molimard)
The European Report ‘‘Lifting the SmokeScreen’’:
Epidemiological study or manipulation?
R. Molimard 1
Faculté de médecine Paris-Sud, 63, avenue Gabriel-Péri, 94576 Le-Kremlin-Bicêtre cedex, France
French version received February 28, 2008 ; accepted June 6, 2008
Available (in French) on the Internet August 13, 2008
Background. – This report ends up with a large increase of the number of estimated deaths from passive smoking in Europe. Its conclusions have been decisive to passing laws banning smoking in public places.
Results. – However, analysis of this report for France reveals considerable anomalies, such as changing the usual definition of passive smoking. Among the 5863 estimated deaths, 4749 concern everyday smokers. The 1114 non-smokers include all the former smokers as well, whose remaining risk cannot be attributed to environmental smoke. Published data used for calculations come from polls, of which results are very diverse and not very reliable. The level of confidence is not discussed.
Conclusion. – Nevertheless, the number of estimated deaths in the field where restrictive laws might be applied and justified is so low that it makes you wonder what the the true aim of this report is, and what conflicts of interest are hidden behind it.
2008 Elsevier Masson SAS. Tous droits réservés.
Up until now tobacco smoke was a nuisance. Many a non-smoker often had the urge to send to hell those who lit up a cigarette at the next table just as their favorite dish was being served. But undeniably, slowly but surely the Evin legislation had brought about change. We were no longer smoking in business meetings, in trains, at the post office, in banks. Slowly but surely, smokers got accustomed to reconsidering before lighting up.
In February 2006, the report "Lifting the Smoke Screen 10 reasons for Smoke Free Europe" was published under the advocacy and recommendations of the European Cancer Society, Cancer Research UK, the European Health Network, and the National Cancer Institute (1) attributing unequalled mortality rates to passive smoke. More than doubling the most pessimistic past estimates and supported by such scientific authorities, it largely contributed to lead the pathway to very strict legislative measures against smokers.
According to this report, such legislations were supported by 75% of the public who had been waiting for them for some time. This is not surprising considering that non-smokers are the majority and a number of smokers regard these coercive measures as the extra incentive they require for the cessation of their habit. However, the announcement of 5863 deaths, quickly rounded off to 6000 in the media by the official tobacco experts, henceforth made environmental tobacco smoke a very serious public health problem. Urgent drastic measures were therefore justified. Finally, a decisive argument was, that although one can stop going to a discotheque or a tobacconist, the employees of these venues, who are forced by present economic hardships to accept such employment, seriously compromise their health.
However, although the majority of the French public benefits from these measures, minimally in the form of comfort, the legislative consequences of this report upset considerably the lifestyles, the culture and the livelihoods of many others. Well aware of the severe tobacco dependence, we can be rightfully concerned about the counterproductive effects in our battle against tobacco, particularly the risk to intensify the smoker’s defensive mechanisms as it pertains to their identity. It is still too early to evaluate the adverse effects and the impact on the benefits/deficiencies ratio of these measures. Consequently, I found it a necessity to analyze the report in an effort to assess the pertinence of its conclusions.
1) ‘’ PASSIVE SMOKING’’ REDEFINED
The terms passive smoking, involuntary smoking or Environmental Tobacco Smoke (ETS), have been defined by consensus to mean the inhalation of another person's smoke by a ‘’non-smoker’’ and its consequences, for over the last 30 years. The report totally changes this definition. The 5863 deaths attributed to passive smoking in France are including the whole population - a large majority of smokers and non-smokers alike. One is to expect that if the air is filled with smoke, of course these active smokers inhale as much if not more smoke as the non-smokers since they do not find it necessary to avoid the company of smokers. They are in effect stricto sensu exposed to environmental tobacco smoke. Inhaling smoke produced by their own tobacco is however inherent to the risks they take as active smokers. To measure the passive ratio of their mortality under the pretense that they do not only inhale their smoke directly but also the smoke laden air of the room in which they smoke, seems artificial and particularly specious. This considerably extended vision of the definition, insidious since it is not discussed, gives good reason to question the objectivity of this report.
The seemingly rigorous methodology is supported by assumptions, that when accumulated, tend to significantly bias the findings. Citing from the report: ‘’… the formula to calculate the ratio of risks related to passive smoking in the population used relative risks for passive smokers published in the recent United Kingdom’s Royal College of Physicians report. They were median values extracted from the data (table III)… It was assumed that ‘’all hospitality industry employees in the European Union’’ were exposed to environmental tobacco smoke in 2002. The estimations used median risks for workplace exposures in the hotel/restaurant sector, but increased risks in the pubs/bar and night club sectors, based on the cotinine levels measured on employees of non-smoking bars…’’ (Table I).
Table I: (From Table 3 of the report) Relative risks associated with passive smoking
Hence, a coefficient of risk increases is applied according to the level of exposure, derived from already published results, for example the risk of non-smokers related to their spouse’s smoking, modulated according to the impregnation of nicotine depending on the place of exposure. This coefficient is applied to the whole population, ‘’smokers and non-smokers alike’’, depending on whether or not they live or work in a place that allows smoking or not. The level and duration of exposure to the smoke being of course unknown, we can already appreciate the level of uncertainty as to the results.
As it is, without a word of warning, without any statistical confidence level allowing for the evaluations of the statements brought forth, the report calculates ‘’to the nearest death’’ the estimated number of deaths. (Table II)
Table II. Estimated annual deaths attributable to passive smoking by age, site of exposure ande condition, France 2002. From Table 6 of the report
This table pertains to the exposure to environmental smoke of the general population. Mortality related to active smoking clearly reflects on the elevated incidence of lung cancer on the younger population. The most recent meta-analysis available confirms the risk of never-smoking women whose husband is a smoker (2), especially that out of 101 studies, only the 55 referring to seven cohort studies or population and control studies were considered. In fact, the reliability of extrapolations from the data available, often derived from questionnaires and surveys, is far from guaranteed. For Europe, it finds a relative risk of 1.31 (CL95 %: 1.24-1.52), but as low as 1.15 (1.03-1.28) for North America. Is it then surprising that passive smokers endure the same kind of risk as current smokers? We can even suppose that the risks are similar in the case of intense exposure. However, they are relatively low in their entirety. The level of exposure is variable and difficult to quantify. The size of apartments for example, can play a determinant role in accounting for the differences between continents. On the other hand, some well conducted studies stand out in the results of the meta studies. For instance, the Europeans multicentric IARC study did not find any evidence of significant harm for neither childhood exposure (OR=0.78; CL95%: 0.64–0.96), nor for spousal exposure (OR = 1.16 ; 0.93–1.44), or workplace exposure : (OR = 1.17 ; 0.94–1.45). For restaurants specifically, no harmful effects were found (OR = 1.03 ; 0.82–1.29). It only found significant effects for intense workplace exposures combined with daily exposures over a long number of years (OR = 2.07 ; 1.33–3.21) (3). Similarly, Hill et al. (4) found no increased risk for lung cancer or respiratory ailments in a new prospective study of two cohorts of never-smokers. Only the incidence of vascular accidents were slightly increased (RR = 1.01 and 1.35, depending on the date and the gender of the cohort. The paradox of this finding is that we would have expected to find that passive smoking would essentially be causing respiratory problems. As it pertains to cardiovascular complications, the findings tend to suggest that a factor other than the direct smoke inhalation comes into play. For instance, smokers tend to make atherogenic food choices that might be shared with their spouses (5). As far as myocardial infarction is concerned, a cohort study found no increased risk in relation to passive smoking (6).
The Office français du tabagisme (OFT) issued a press release in March 2008, to communicate the findings derived from the Oscour data of the Institut de Veille-Sanitaire (InVS). They reported that the first seven weeks of smoking bans in the hospitality industry beginning January 2 resulted in a 15% reduction of hospital emergency admissions for myocardial infarction (MI) (7). The MI incidence is expressed in percentages for 100 000 admissions and therefore sensitive to variations in other causes. Data on the absolute number of MI incidence admissions are absent. We are unable to go back to the sources of the data published in the OFT website. It is therefore impossible to argue extensively except for pointing out that similar decreases have already been observed several times in the past two years. But the contradiction is in any event obvious when it comes to the report that was used to justify such prohibitions, because the estimate of annual deaths occurring in non-smokers in this environment works out to six deaths all causes combined. Even if similar results were published elsewhere (Italy, U.S.A., U.K.), caution should have been used before concluding publicly to a causal relation.
2. NON-SMOKERS AND FORMER SMOKERS
The report eventually gets to address what it is allegedly all about - the non-smokers. To calculate their number, they first calculated the number of smokers, by multiplying for each country the total population by the percentage of its smokers. The number of non-smokers is obtained by simply subtracting from the total population the number of smokers thus calculated.
The estimation of the number of smokers is, as we can see, central in this study. But it’s a totally arbitrary figure. No figure is indicated for France. The only figures utilized that allegedly represent Europe, are those of Italy (26.6%), Slovakia (28%), Spain (28.1%) and the United Kingdom (26.8%). For the remaining countries, they applied percentages of other countries, either because they’re neighbors, (Italy and Greece, Spain and Portugal), or ‘’because they speak the same language’’ (sic!) for the United Kingdom and Ireland. These figures apparently come from the WHO (8), but why then did they not use the available figures instead of using ‘’assimilations’’? It gives the impression that the work was rushed off. Worthy of mention is that the data of the WHO must be relativized because they are in total contradiction with the ‘’Eurobarometer’’ survey in 2002, that did not find in Europe any prevalence inferior to 35% and went as far as finding 45% prevalence rates in the U.K. ! (9). All these percentages are obtained through polls, house to house and telephone surveys, without any biological controls. The level of uncertainty is therefore considerable. The report does not reveal to which country France was assimilated. Yet we do have data for France. For the same year, the WHO was reporting 27% smoking prevalence, a INPES-IPSOS survey 30.4% for the 15 – 75 year bracket (10), the CSA Institute for the Eurobarometer 44%! This is very revealing of the nature of the foundations this report was built on!
The method used for calculating the number of non-smokers thus includes all never-smokers as well as the former smokers. And the report gives the following precision: ‘’…All excess risk of their past smoking status was ignored’’ for three reasons :
- Data from the WHO does not segregate between ‘’never smokers’’ and ‘’former smokers’’;
- The excess risks associated with past smoking decreases with time, but there is very little documentation to correctly define how this decrease occurs in relation to the four major illnesses. Most refer to studies conducted when non-filtered cigarettes occupied an important part of the market (sic!);
- Even if we could precisely know how this decrease occurs, applying this knowledge would have required detailed information on the date of the smoking cessation for each individual, information that is not known.
The argument that the WHO does not segregate between former and never smokers is inadmissible. Such data do in fact exist. For France, an INPES survey in December 2003 found 38.4% never smokers and 31.1% former smokers (11). Spreading these results to the national population could have been questionable, but 45% of former smokers in an alleged non-smoker group is an order of magnitude that would have at least commanded an analysis of the bias that it possibly created.
However, after having stopped smoking, one does not from one day to the next retrograde to the same risk levels as one who has never smoked. Indeed the risks decrease with time depending on the duration of the abstinence. But the famous Doll study on British doctors found that for lung cancer the RR is 16 between one and four years of cessation, 5.9 between five and nine years. It remains however at twice the level of a never smoker even after 15 years! (12). The excess risk for MI is still detectable after ten years of abstinence (13) and the harm done to the bronchi is irreparable. The way the report is designed, it can very well classify as a non-smoker anyone who has smoked for 30 years and stopped 15 days ago! Passive smoking plays absolutely no part in these belated consequences of active smoking. Ignoring the excess risks of the former smokers is therefore unacceptable. It is baffling how scientists can dismiss such selection bias that can completely distort the results. As for the non-filter cigarettes comment, it would bring a smile had it not been used in an effort to suggest that filters can decrease the residual risks to the point of insignificance. We unfortunately know that this is not the case at all! (Table III)
Table III Estimated annual deaths of non-smokers attributable to passive smoking by age, site of exposure and condition, France 2002. From Table 8 of the report
Deaths of non-smokers as defined by the report are now down to 1114 deaths. Fortunately however, the legislation does not apply to private home smoking yet, although the Association for Smoking and Health (ASH) is beginning to lobby at least the U.S. government for such measures. It would therefore apply to the 107 annual deaths corresponding to workplace exposure only, of which a mere six pertain to the hospitality sector.
Thus, the determinant point to justify the extreme severity of the legislation prohibiting smoking, even in places where tobacco is sold, is the health risks for the workers of such places. During a visit in Sens on January 18, 2008, the President of the Republic replied to a tobacconist: ‘’There is no freedom for those who enter a smoke laden place. The employees and the non-smoking consumers are not obligated to contract a cancer because others have so chosen, to the extent of how capable to make such a choice 16 or 18 year old youngsters are.’’
According to the INPES data, 45% of the 107 deaths would be comprised of former smokers. When applying Doll’s findings, we can estimate a median residual RR of 6 for lung cancer, that would account for one third of deaths (12), and 1.5 for the other ailments (13). It therefore represents an excessive risk factor of 3 to be applied to this group, representing a real measure of approximately 70%. We would therefore now have around 30 annual deaths attributable to passive smoking in the workplace, thus 2 deaths for the whole hospitality industry sector! What fraction of the mortality risk can then be attributed to smoking at the bar-tobacconists that would justify the overkill legislation that hit them?
Even if we only account for the total figures in the report, their order of magnitude is far off from what would be considered a serious public health problem that would justify the advocacy of such repressive legislations. The intention is clearly expressed in the title: 10 reasons for a smokefree Europe. It becomes therefore legitimate that we question the true motives of the legislation. The answer may well be found in the introductory chapter where one can read: " …as the response to the Smoke free Europe Conference on 2 June 2005 revealed. Held in Luxembourg under the auspices of the Luxembourg Presidency of the EU, and organised by the commissioning organizations of this report with sponsorship from GlaxoSmithKline and Pfizer, the conference brought together at European level for the first time health organisations,leading researchers and representatives of Europe’s public and private sector employers, trades unions, occupational health inspectors, the European Commission and politicians to debate smoke free policy. In total, nine serving ministers of health and/or employment plus the ex-health minister of Italy spoke at the event.’’
Wasn’t Europe capable of financing the debates of such an important issue? Is it not scandalous that a European Conference organized by those who wrote the report, at which ministers and the European Commission participated, was sponsored by two international giants of the pharmaceutical industry, who market smoking cessation products, Niquitin* and Zyban* (GSK), Nicorette* and Champix* (Pfizer), that have succeeded to obtain coverage by medical insurance (50 EU a year per smoker), all the while continuing to market over the counter medication? Their lobby groups are very active and efficient indeed! Wouldn’t this basically constitute an indignant manipulation? A report that cares very little for scientific integrity doesn’t make one too confident as to the way subsequent criticisms will be treated.
The whole industry’s morality is profit oriented. The tobacco industry’s deviations have been rightly denounced and condemned since decades. Curiously, the pharmaceutical industry, has changed its denomination to the more flattering ‘’Entreprises du médicament‘’ (The Medication Companies) and now benefits from a suggestive angelic aura and the noble image of some savior prince on his white horse. Clearly their lobby groups infiltrate decision makers at many levels. Is it not high time we started restraining their behavior?
Trampling on the very elementary scientific principles, this report can only discredit the cause it claims to defend. The end does not justify the means. Fighting tobacco dependency is a public health necessity. But we have to learn from the lessons of past experiences. Extreme repressive policies, whether the excommunication (Urban VIII) or the decapitation and the hanging (Amurat IV), nothing succeeded to restrain the worldwide spreading of the use of tobacco. (14,15). We should justifiably worry that purely restrictive laws, advocated on biased bases, are going to set off the identity defensive mechanisms of smokers which would only result in the exact opposite direction of the official goals we are striving to attain, as explained in the experimental psychology studies (16, 17). Such policies tend to make us lose focus of the reality that it is the smoker who is the true victim of tobacco and that he deserves more attention than hostility. This attention must be canalized through serious scientific studies on tobacco and its dependency and must take into consideration not only the epidemiological and neuropharmacological aspects of the issue, but the psychological and sociological as well. The enormous amounts of money at stake would dictate that it must be done in complete independence of lobbies, not only the tobacco industry’s but the pharmaceutical industry’s alike.
4.- CONFLICTS OF INTERESTS
According to article R. 4113-110 of the 2007-454 implementation decree March 25, 2007 article 26 of the March 4, 2005 law (L 4113-13 article of the Public Health Code), which sets up the transparency of the medical information, I declare on my honour that I do not benefit of any material or moral advantages from neither pharmaceutical firms, nor the tobacco industry
 « Lifting the smokescreen, 10 reasons for a smokefree Europe ».
 Taylor R, Najafi F, Dobson A. Meta-analysis of studies of passive smoking
and lung cancer: effect of study type and continent. Int J Epidemiol
 Boffetta P, Agudo A, Ahrens W, Benhamou E, Benhamou S, Darby SC,
et al. Multicenter case-control study of exposure to environmental
tobacco smoke and lung cancer in Europe. J Natl Cancer Inst
 Hill SE, Blakely T, Kawachi I, Woodward A. Mortality among lifelong nonsmokers
exposed to second-hand smoke at home: cohort data and
sensitivity analyses. Am J Epidemiol 2007;165(5):530–40.
 Lepeltier D. Concealed causes of lipid consumption by smokers: nicotine and proteins (in French) Medicine thesis. Faculté de médecine Xavier-Bichat - Université Paris-VII, 1994.
 Stranges S, Matthew R, Bonner MR, Fucci F, Cummings KM, Freudenheim
JL, et al. Lifetime cumulative exposure to secondhand smoke and risk of
myocardial infarction in never smokers: results from the Western New
York Health Study, 1995–2001. Arch Intern Med 2006;166:1961–7.
 WHO Data. http://data.euro.who.int/tobacco/
 Eurobarometer. http://europa.eu.int/comm/public_opinion/archives/eb/
 INPES study. www.inpes.sante.fr/30000/pdf/0402_synthese_tabac.pdf (in French)
 INPES. Actualités Tabac. 2004, n847. www.inpes.sante.fr/TA/pdf/
Tabactu47.pdf (in French)
 The health consequences of smoking. Cancer: a report of the surgeon
general. US department of health and human services, 1982: p.46.
 Quia Q, Teruahanta M, Niessen A, Tuomileho T. Mortality from all causes
and from coronary heart disease related to smoking and changes in
smoking during a 35 year following of middle aged Finish men. Eur Heart J
 Chaudon LM, Delandine AF, Brotier G, Mercier B. Dictionnaire universel,
historique, critique, et bibliographique 1811, p 385. http://
books.google.com/books? (in French)
 Encyclopédie de Diderot et d’Alembert. http://diderot.alembert.free.fr/
N.html (in French)
 Falomir JM, Mugny G. Society against smoker: a psychological analysis of the experts's influence (in French)
. Grenoble: Presses Universitaires de Grenoble;2004.
 Falomir JM, Mugny G. Identity stakes and resistance to anti-smoking campaigns. (in French) http://tabacologie.fr, Cours du DIU n836