Tobacco Documents Online

~ulticomponent "Lea.Kue Tables" S.J.G. 8 SEPI~]] nL~,~ ....... I . Within the past two years, an undesirable development of the traditional type of "League Table" has taken place in certain countries, in that they have been expanded from the simple tables of deliveries of "tar" and nicotine to include one or more components of the gas phase of cigarette smoke? The commonest gases to be measured have been carbon monoxide and the mixed oxides of nitrogen, but there have been listings of deliveries of hydrogen cyanide (Reader's Digest - US brands) and suggestions to measure ammonia (Netherlands), acetaldehyde, hydrogen cyanide and acrolein (Switzerland). The advent of multicomponent "league tables" has raised problems in ease of comprehension by consumers, who findit difficult to assess the relative importance of the different factors. The undesirable development is the ways in which different publications have amalg,~mated the data into rating indices for the various brands and this is best described country by country. Germany The first multicomponent league table was published in "Test", the magazine of the consumer foundation St~ftung Warentest in March 1975. This reported data on dry particulate matter (DPM), nicotine, carbon monoxide and oxides of nitrogen for 37.brands, but did not attempt to produce an Index figure combining the various analyses. Switzerland (a) The Swiss Consumer Protection Foundation in February, 1976, published data on the same four smoke components ~rom 31 Swiss brands in an article "Test 73" and, on this occasion, combined the figures into an index of Points out of I00. The method whereb~ this Herzfeld Index (Dr. Herzfeld devised the Index) was calculated, was comp]icated and laborious. It is not described here as it has been ~uperseded by:- (b) Herzfeld Index Mark II, published in April, 1977, as "Test i01", covering 44 Swiss brands and measuring DP~I, nicotine, carbon monoxide and oxides of nitrogen once again. For each brand, the del~.veries of each smoke component were expressed as a percentage of the deliveries of a fictitious cigarette "yielding":- DPM 30 mg Nicotine 2 mg Carbon monoxide 6% v/v O~idos of nitrogen1500 ppm BAT Industries document for Province of British Columbia 13 November 2000 BAT INDUSTRIES 00290406

-2- The values were then added and the sum expressed as a percentage relative to 300 (not 400, as might have been expected on a logical basis). The gamut of Indices between 0 and 100% was then divided into five categories:- 0-20% Extremely mild 1 open circle 20-40% Mild to aromatic 1 dot 40-60% .Aromatic to strong 2 dots 60-80% Strong 3 dots 80-100% Very strong 4 dots It was suggested thdt the category of the brand could be designated on the pack by visual marks ranging from an open circle to four black dots. In March 1977, the Belgian Consumer Magazine "Test-Achats" published analyses by the Dutch TNO Laboratory for 41 brands sold in Belgium. On this occasion, analyses were made for mainstream DPM, nicotine and carbon monoxide and analysis for oxides of nitrogen was not undertaken. Instead, a figure was given for the amount of sidestream carbon monoxide (in mg per cigarette). For each smoke component, the range of deliveries was divided into four categories, indicated by one to four dots:- NS. of dots DPM Nicotine Mainstream 8idestream carbon monoxide carbon monoxide 616 gO. 8 ~ 19. ~53 • . 17-20 0.9-1.1 13-14 54:59 • . . 21-25 1.2-1.4 15-17 ~60-67 ..... ~.26 ~.1.5 ~.18 -" ~68 Then, for each brand, the number of dots for each smoke component is totalled and the brand is then allocated to one of three categories:- Up to 7 dots 8-9 dots I0 dots or more Less noxious than average Average More noxious than average BAT Industries document for Province of British Columbia 13 November 2000 BAT INDUSTRIES 00290407

-3- United States In late 1976, "Reader's Digest" published a two-part article on the gas phase of cigarette smoke, listing the deliveries (in mg per cigarette) of carbon monoxide, oxides of nitrogen and hydrogen cyanide. A combined "triple-gas rating" for 28 US brands was produced by simply adding together the tllree deliveries. ~ Th~ "Reader's Digest" articles were also noteworthy for recording delivery data obtained under smoking conditions different from the standard puff volume and frequency, attempting to mimic a range of human smoking behaviour. Discussion None of the proposed methods of aggregating the various measures of smoke deliveries is soundly based Jn scientific terms. Implicitly, they assume that the alleged "hazards" associated with the various smoke components can be accorded equal weight, e.g. that the risk to the smoker from inhaling nicotine in milligram quantities may be equated to that incurred by inhaling an atmosphere containing a few percentage parts of carbon monoxi4e. Indeed, the first attempt by Herzfeld to establish an Index rating was accompanied by a commentary by a well-known Swiss doctor, which implied that each component was specifically related to different diseases which have been associated statistically with the smoking habit, e.,g. DPM with lung cancer, carbon monoxide with cardiovascular disease, oxides of nitrogen with emphysema. Such specific relationships are clearly untenable. Non-smokers succumb to cardiovascular disease, the etiology of which is universally accepted as multi-causat2ve, allegedly embracing physiologic type, life style, diet, etc. as well as smoking habit. Similarly, other smoking-associated diseases have other possible causes besides smoking. To single out specJ.fic smoke components and relate these to specific diseases is even less plausible. In Switzerland, the Director of the Federal [iealth Department stated publicly that it was '~not possible to combine four different poisons in a single list". Nevertheless, such aggregation methods have been recognised by anti-smokers as a potent weapon to persuade manufacturers to design products with ever-lower deliveries, with the probable latent to wean smokers from the habit. BAT Industries document for Province of British Columbia 13 November 2000 BAT INDUSTRIES 00290408

-4- Conclusion Multicomponent league tables confront the Industry with a dilemma. Further developments in cigarette brand analyses, embracing additional smoke components, can be foreseen an~, faced with ever more complex tnbles, the consumer may be expected to react by demanding some kind of simplification such as an extension of existing methods for calculating Index Ratings.- Rejection of this demand with the argument outlined above that these methods are based on unsound foundations will lead to a request for a scientifically valid way of combining data. This cannot be undertaken without proposing weighting factors for the various smoke components, based on the strength of the statistical association of smoking with different diseases and an estimate of the likely etiology of a specific disease, thereby tacitly admitting a degree of causality - an unacceptable conclusion for legal reasons, even if the requisite data were available. On the other hand, marketing departments w~sh to ensure that their own brands mre well-placed vi~ avis the competition when multicomponent tables are published and, as a result, wish to know how the brands ~nd new product developments rate using, for example, a ]lerzfeld Index system. Thereby, they give a degree of credence to a method of evaluation which, in the lon~ term interests of the industry, should be discredited. The Company Bomrd Plan is quite unequivocal on this point:- "Such arbitrary indices of hazard cannot receive any serious sc~.entific support and their introduction should be resisted wherever possible". The preferred course of mction, when threatened with the introduction of such indexin~ systems is to involve the Industry organisation and, through them, to present the. Government authorities with areasoned argument in refutation. In the U.K., where the imminent publication of tables of carbon monoxide deliveries can be foreseen, the assistance of other companies has already been sought through the Tobacco Research Couhgil. The intention is that this body will make representations to the Dep~rtment of Hemlth through the Independent Scientific Advisory ..Committee (the "Hunter Co,~ittee") in the expectation that an authoritative pronouncement would be made against any attempt to publish an aggregated "hazard" rating. The value of a Government- --- statement, such as that made in Switzerland, noted earlier, can be considerable. DGF/AII~/1.1.3.1 19th August, 1977 BAT Industries document for Province of British Columbia 13 November 2000 BAT INDUSTRIES 00290409