LZ~' AW 139 0 9, W-0 " ~ 44~ I~D _1~j tl%i BATCo document for Legal Services : Health Canada 20 October 1999 I 000001, P.. logo, CPNPD M Ib 1965 RE 71oo 0 1~ej4 C) \10 BATCo document for Legal Services : Health Canada 20 October 1999 Sam hnperial Tobacco Limited/Limitee 13 0 RESEARCH AND DEV LOPM NT DIVI ION MONTREAL THIS REPORT MUST NOT BE COPIED OR SHOWN TO UNAUTHORIZED PERSONS. ,,Io BATCo document for Legal Services : Health Canada 20 October 1999 Research Laboratory Report f_174 Copy No. -3- IMPERIAL TOBACCO LTD. RESEARCH & DEVELOPMENT DIVISION MONTREAL RESTRICTED P- ~ I V%,Y% 's INVESTIGATION OF THE INTERACTIONS OF SMOKER BEHAVIOUR AND CIGARETTE DESIGN AND THEIR INFLUENCE ON DELIVERY AUTHOR: C. McBride ISSUED BY: S.R. Massey DATE 155UED:February 12, 1985 Distribution Dr. P.J. Dunn Copy No. 1 Dr. S.R. Massey 2 Mr. A.L. Heard 3, 4 Mr. E.E. Kohnhorst 5, 6 Mr. R.G. Nicholls 7 Herr. E. Rittershaus 8 Dr. F. SeehoW 9 Dr. C.J.P. de Siqueira 10 Library 11. 12, 13 N) 4~ C.D BATCo document for Legal Services : Health Canada 20 October 1999 TABLE OF CONTENTS SUMMARY INTRODUCTION ACKNOWLEDGMENTS 1. DISCUSSION 1 .1 Comparison of Human Smoking versus Standard Machine Smokinq Deliveries 1.2 Carbon Monoxide 1.3 Comparison of Cicarette 919 versus Matinde Special Filter 1.4 Comparison of Standard and Human Per Puff Deliveries 1.5 5ublective Impressions 1.6 Comparison of Per Puff Parameter Changes 1.7 Comparison of Delivery Chanqes an a_Per Puff Basis 1.8 Comparison of Human Deliveries and T/N Ratios 1.9 ComDarison of ObJective Smoking Parameters and Sublective Resnonses 2. CONCLUSIONS 3. APPENDICES 0 1 Methods Il Subjective Questionnaire III Calculation for the Mean Parameter Changes and Frequency of :hange Terms. -C>. CD r\) _t:~k "0 -,4 CK BATCo document for Legal Services : Health Canada 20 October 1999 SUFMARY This study investigated the Interaction between cigare:-.e design and smoker behaviour and their Influence on delivery. This was acc:mplIshed by assessing the Influence of smoker behaviour on both the total ar'd puff-by-puff de'lveries of two cigarettes of similar total tar delivery but tifferent subjective characteristics. Discussed in this report is a method for the determina-.1on of smoking behaviour changes on a per puff basis, their effect on cigarette delivery and the Influence on smoker preference. In general, it was found that: 1) although the human mouth deliveries were greater than the deliveri es un standard machine conditions, human T:N and CO:T ratios were in close agreement with machine delivery ratios. 2) there was a tre nd for human mouth deliveries and smoking parameters to decrease as the cigarette was smoked. 3) behavioural changes during a switching exercise were more evident on a per puff basis than an a total cigarette basis. 4) the degree to which smokers changed their behaviour and the Jegree to which cigarette delivery was affected by these changes and z-:)oeared to Influence product acceptability. 5) certain subjective characteristics such as irritation and draw effort can be associated with the objective smoking parameters - flow rate, interval and per puff T/N ratios. KEYWORDS: SMOKER-PRODUCT INTERACTIONS HUMAN SMOKING BEHAVIOUR SUBJECTIVE-OBJECTIVE RELATIONSHIPS ~10 I-J BATCo document for Legal Services: Health Canada 20 October 1999 - 2 - INTRODUCTION A previous study (unreported; In-House Brand Survey, August. 1981) suggested that smokers modify their smoking behaviour during smoking to satisfy personal requirements. It became apparent that the smokers' perceptions of a product were influencing their manner of smoking and ultimately their assessment of and preference for a particular cigarette. The objective of this study was to assess the influence of smoking behaviour on both the total and puff-by-puff deliveries of two cigarettes which had similar total tar deliveries under standard machine conditions but different subjective characteristics. As noted above there was evidence to suggest that smoking behaviour was modified during smoking i.e. on a puff-by-puff basis. Therefore it was felt important to investigate both total and puff-by-puff behavior and delivery. This was a preliminary study to investigate the interaction of the smoker and the product in terms of the smoker's perception of the cigarette, and smoking behaviour, and'the delivery of the cigarette. The cigarettes selected for this study were a non-ventilated brand and its ventilated equivalent. Both cigarettes were designed to deliver approximately 9 mg of tar under standard conditions but with different subjective characteristics. These two points are particularly relevant with respect to the direction of the Canadian market towards lower delivery, highly ventilated brands. This report discusses a method for assessing the behavioural changes smokers undergo while smoking, the effect of these changes on the human delivery of a product and how this may influence a smoker's preference. ACKNOWLEDGEMENTS The technical assistance of Mme R. Corbin and M. G. Prifontaine Is gratefully a cknowledged. 1t~b CD CD "ci -1J Cr.) BATCo document for Legal Services : Health Canada 20 October 1999 - 3 - 1. MISCUSSION Details of the cigarettes are given in Table I- The two products delivered approximately the same tar (9 mg) and nicotine (0.8 mg) under standard conditions but were perceived to be subjectively different. This was achieved through the use of a high efficiency non-ventilated filter (SCS III) on the control (Matinde Special Filter) and a high pressure drop low efficiency filter (Filtrona CDD) which was 34% ventilated on the experimental cigarette (919). 1.1 Comparison of Human SmokiAg__versus Standard Machine Smok1nQ Parameters and Tar and Nicotine Deliveries For a majority of smokers, mouth deliveries were greater than 40 standard machine deliveries (Table II). These increases are due to the differences between the smoking parameters of humans and standard smoking machines, which are given in detail in Table III. The main points can be summarized as follows: i) Relative to machine smoking the subjects drew a larger total volume of smoke into their mouths. ii) The increase in total volume Is related to an increase in individual puff volume and puff number (i.e. decrease in puff duration and interval). iii) The average puff flowrates for human smokers was significantly higher than for standard smoking conditions. The last point is of particular importance since the efficiency of the filter decreases as smoke velocity increases. in general, the human velocities were 68% higher than standard machine velocities (Table III). Thus the filtration efficiency of both cigarettes was lower at the velocities generated under human smoking conditions than under standard machine conditions. In addition, for a filter ventilated product, increases in puff r~J 4--b CD \10 BATCo document for Legal Services : Health Canada 20 October 1999 - 4 - velocity tend to give rise to a reduction in the proportion of diluting air entering via the filter. It should be noted. however, that although changes in behaviour affected the absolute amounts of individual smoke components drawn into the mouth, the average human tar: nicotine ratios were in close agreement with those determined under standard conditions (Table II). 1.2 Carbon Monoxide Relative to the standard machine delivery of carbon monoxide, the average human delivery of carbon monoxide for cigarette 919 was 42% higher. In the case of Matin6e Special Filter (MSF), the smokers showed only a 6% increase In carbon monoxide relative to standard machine smoking. ICARBON MONOXIDE DELIVERY (mg/ciqt.)l MACHINE HUMAN MATINU S.F.1 16.2 17.2 1 CIGT. 919 1 9.2 13.1 1 However, on average, the subjects were exposed to 23% less carbon monoxide in their mouths with cigarette 919 as compared with MSF. It is apparent, then, that cigarette 919 which was designed to give a reduced CO: tar ratio under standard machine conditions does result in smokers being exposed to lower absolute levels of carbon monoxide and a reduced CO: tar ratio. The relative inhalation factor for each subject (Table IV) indicates that all but three inhaled less of cigarette 919 than MSF. With cigarette 919, subjects had a higher ratio of available tar and nicotine to CO In their mouths than with MSF (Table II). Thus, to obtain the same amount of tar and nicotine In their lungs, it was not necessary for subjects to inhale to the same extent with cigarette 919 as with MSF. -t~> CD rv (ZD "0 cc CD BATCo document for Legal Services : Health Canada 20 October 1999 - 5 - 1.3 ComDarison of the Human Smoking Parameters and Deliveries for Cigarette 919 versus Matin6e Special Filter The data In Tables 11 and III show that on switching from MSF to cigarette 919, smokers were able, with relatively small changes In smoking behaviour to increase their mouth deliveries of tar and nicotine, by 22% and 35% respectively. These increases were not. however, significantly different from what would be anticipated, given the differences in the standard deliveries of the two cigarettes. With cigarette 919, smokers took, on average, one more puff than with MSF. It seems likely, that this is the major reason for higher total mouth deliveries with cigarette 919. Although the average differences in total delivery between MSF and cigarette 919 were not unexpected, it is interesting to compare the population distribution of mouth deliveries for the two cigarettes relative to their deliveries under standard conditions. For each cigarette, the population of smokers was divided into 3 groups: those with mouth deliveries within + 10% of machine deliveries, those with mouth deliveries less than 10% below machine deliveries and those with mouth deliveries greater than 10% above machine deliveries (Fig. 1). For MSF, the characteristics of the groups were as follows: 1) 44% of the smokers achieved mouth deliveries within a range of ñ M* of machine deliveries 2) 25% achieved mouth deliveries considerably below this range (average of 4D% less than machine deliveries) 3) 31% achieved mouth deliveries above this range (average of 50% higher) * + 10% of machine deliveries was thought to be indicative of the range In delivery which could be a attributed solely to product variability. CD \.O BATCo document for Legal Services : Health Canada 20 October 1999 - 6 - In contrast, with cigarette 919. the groups could be characterized as f ol lows: 1) 19% within + 10% of machine deliveries 2) 25% below (average of 18% less) and 3) 56% above (average of 51% more) Some possible reasons for the increase in human mouth deliveries relative to machine deliveries as well as the differences between the mouth deliveries for the two cigarettes, have been discussed previously in thIs report. The data in Figure 1. not only illustrates these differences but also the smoker-cigarette interactions of the two products. For the purposes of this study, smoker- c i garette interaction was defined as the relationship between the smoker and the cigarette in terms of the smoking behaviour of the subject and the delivery of the cigarette. A previous study (unreported; In-House Brand Survey, August, 1981) suggested that: 1) smokers modify their behaviour during smoking and 2) smokers of ventilated products do so to a greater extent than do smokers of non-ventilated-brands. It is apparent from Figure 1, that the smokers in this study interacted differently with '-he ventilated cigarette (919) as compared with the non-ventilated ciSarette (MSF). 1.4 Comparison of Standard and Human Per Puff Deliveries The per puff deliveries obtained under human and machine Conditions were examined. This indicated that the human per puff profiles -were quite different from those obtained under standard machine conditions (Figures 2, 3, 4 and 5). The variable nature of human smoking behaviour is emphasized by the comparison of these per puff delivery profiles. This supports t~e idea that smokers are modifying their behaviour while smoking. It also inolies that 1t:2. CD Co r~J BATCo document for Legal Services : Health Canada 20 October 1999 7 smokers must be receiving feedback from the cigarette in order to do so. Therefore the definition of smoker-cigarette interaction must also include a measure of a smoker's perception of the cigarette. Thus, to investigate smoker-cigarette Interaction, it is necessary to assess: 1) the smokers' specific responses to each product in terms of subjective evaluation and smoking behaviour and 2) each product's response, in terms of delivery, to the smokers. 1.5 SubJective Impressions To have an appreciation of the smokers' perception of the two cigarettes, they were asked to complete a subjective impression questionnaire (Appendix II). From the data in Table V, it can be seen that the two is cigarettes were significantly different in all aspects of subjective evaluation except draw effort. Since they were selected to similar pressure drop levels (Table I), the similarity of draw effort was not unexpected. On the basis of a forced choice, the population was divided into two preference groups (Table Va). Further work in this study was based on the similarities and dissimilarities of these two groups. Specifically, the per puff behaviour and delivery differences between the two groups were compared when the smokers swi tched, f rom the non-venti 1 ated to the ventl lated product. The relative per puff differences between cigarettes, as used in this study, are important for the following reasons: 1) smokers base their judgment of a product both on their previous smoking experiences (in this case the non-ventilated product) and on their experience while smoking the novel product (the ventilated cigarette). 2) data based on total cigarette behaviour and delivery (Tables II and III) did not show the subtle changes and possible cues smokers are experiencing while smoking a cigarette. _c~b- CD CD 1~0 BATCo document for Legal Services : Health Canada 20 October 1999 - 8 - Definitional differences existed among the smokers (untrained In subjective assessment) which prevented a detailed study of the relationships between individual smoking parameters (puff volume, flow, duration and interval) and subjective characteristics (irritation, mouthfull, draw effort, etc.). However, trends were observed regarding some of the interrelationships and these are discussed later in this report. A more detailed study will be the subject of future projects. 1.6 Comparison of Per Puff Parameter Changes The per puff behavioural changes associated with switching from the non-ventilated to the ventilated cigarette can be seen in Table VI. They are also illustrated for some subjects In Figures 6-8. The main points can be summarized as follows: 1) The term 'Change" is a measure of the magnitude of the differences in the puffing parameters associated with the two cigarettes. Details of the calculations are given In the Appendix III. From this, it can be seen that the differences In the behaviour associated with the two products were greater for the group who preferred MSF than for the group who preferred 919. As an example: the MSF preference group increased their volume by 22.9% when switching from MSF to 919, while the 919 preference group showed only a 9.5% increase in volume. 2) "Frequency" Is an Indication of how often the behaviour associated with cigarette 919 was different from the behaviour associated with MSF; details are given in Appendix III. The greater the Frequency term the more often behaviour was changed. It is important to note that these terms are meaningful only in their relative differences and not in absolute values. Again the group who preferred MSF exhibited more frequent differences in behaviour relative to the group who preferred cigarette 919. 3) It is possible to place the various puffing parameters in a hierarchy based on the smokers' apparent willingness to alter them. If the Frequency term in Table VI Is examined, it appears that inter-puff interval is the first parameter to be changed while duration Is very rarely altered. _t!tl c::) co 41- BATCo document for Legal Services : Health Canada 20 October 1999 - 9 - 4) The most radical changes in behaviour occurred in the latter part of the cigarette, usually after the third to fifth puff. This is not illustrated in the table but was observed during Its compilation. It is particularly important to note that when switching, the group who preferred MSF to cigarette 919 changed their behaviour to a greater extent and more often than did the other group, In terms of smoker-cigarette Interaction, the former group of smokers were Interacting differently with the cigarettes relative to the group who preferred 919. It also suggests that the group who preferred MSF were experiencing greater difficulty in satisfying their subjective requirements with cigare.tte 919. 1.7 Comparison of Delivery Changes on a Per Puff Basis Throughout this study, it has been assumed that smokers attempt to make a new product respond In a way that *is familiar to them. A measure of the behavioural changes a smoker undergoes when switching from a conventional .(or familiar) to a novel product reflects the amount of effort required by the smoker to switch. A successful switching exercise, i.e. the smoker prefers the new product to the old, would require that these behavioural changes be minimal and/or offset by a more favourable (i.e. familiar) cigarette response. It is necessary, therefore, to assess the magnitude of the differences between a new product and the familiar product in terms of smoking behaviour (Table VI) and also in terms of delivery to the smoker. Previously in this report, it was noted that the total cigarette deliveries for both products were very similar. However, it can be seen from Table VII that the human per puff deliveries Offered for the two preference groups. The per puff average changes in human deliveries have been normalized with respect to the per puff deliveries obtained under standard machine conditions. Therefore, these figures reflect the differences between the two cigarettes that can be attributed to differences in human smoking behaviour. _t_~ CD CD \IC CD (il BATCo document for Legal Services.: Health Canada 20 October 1999 - 10 - The group who preferred MSF showed the following characteristics relative to the group who preferred cigarette 919: 1) The differences between their mouth deliveries with cigarette 919 and their mouth deliveries with MSF were greater. The group who preferred MSF obtained 23.8% more tar, 26.5% more nicotine and 13.8% less CO with cigt. 919 while the group who preferred cigarette 919 obtained 13.8% more tar, 33% more nicotine and 17.5% less CO with ci9t. 919 relative to MSF. This suggests that for the former group of smokers, the gap between their familiar or usual delivery was wider as compared with the latter group of smokers. 2) The difference between MSF and cigarette 919 fluctuated to a greater extent. This can be seen by reference to the Frequency term which is comparable to the Frequency term in Table V1. The group who preferred MSF showed much higher Frequency terms for all deliveries relative to the group who preferred cigarette 919. Therefore the former group appeared to be altering their behaviour and delivery more often during smoking than the latter group of smokers. The measure of fluctuation in human deliveries during smoking is perhaps an indication of the degree to which: 1) the design of a cigarette is sensitive to variations In human smoking behaviour and 2) smokers are controlling the delivery of a cigarette. The results of this study suggest that a smoker's subjective assessment during smoking allows them to influence a cigarette's delivery, through modifications In their behaviour, in such a way as to meet their individual subjective requirements. The extent to which they can successfully achieve this influences product acceptability. This is evident from the data in Tables V, VI and VII which show that the group of smokers who rejected cigarette 919 felt it necessary to change their behaviour to a greater degree and yet experienced an unsatisfactory cigarette delivery relative to the group who preferred cigarette 919. 1 N) CD Co CY\ BATCo document for Legal Services : Health Canada 20 October 1999 - il - 1.8 Comparison of Human Deliveries and T/N Ratios To clarify the differences between the two preference groups, the individual human deliveries and delivery ratios were examined more closely. The variances of the per puff tar and nicotine deliveries and T/N ratios were calculated. These are presented in Table VIII. The T/N ratios for three subjects are shown in Figure 9. The data indicate that.- 1) for the majority of subjects (10/16) there is a tendency for the more variable delivery, especially tar, to be associated with the preferred cigarette and 2) for less variable T/N ratios to be associated with the preferred cigarette (13 of the 16 subjects) This supports the view that smokers may not be looking for a level delivery cigarette in terms of isolated tar or nicotine deliveries but that the balance of a cigarette's delivery Influences product acceptability. The variability of the individual deliveries reflects the effect of the smokers' modifications to their behaviour while smoking. A balanced combination of smoke components, as indicated by the less variable T/N ratios, may be part of a smoker's desired goal. If such is the case, the group of smokers who preferred MSF did so, perhaps, because they were able to achieve a more level or less variable T/N ratio with MSF as compared with cigarette 919. Future studies are planned to further investigate: 1) whether smokers are changing their behaviour to achieve a balanced cigarette delivery which can be monitored through per puff T/N ratios, 2) whether this balance can be quantified in terms of an acceptable 0 degree of variability in T/N ratios, and 3) how they are related to standard condition T/N ratios. -J~L- CD N> CD OD ---j BATCo document for Legal Services : Health Canada 20 October 1999 - 12 - 1.9 Comparison of Objective Smoking Parameters and Subjective Responses Up to this point. this report has discussed the broad similarities and dissimilarities in the, smoking behaviour and deliveries of the two preference groups. It has been noted that smoking behaviour is a means by which smokers may alter a cigarette's delivery. In turn, this will Influence their subjective assessment and perhaps their overall preference. Therefore, to allow more efficient control over the smoke being delivered to the smoker, it is important to relate specific objective parameters to specific subjective characteristics. Changes occurring in volume, flow, duration, interval and TIN ratios during smoking were compared to the subjects' Impressions of Irritation, draw effort and mouthfull. These comparisons are listed In Table IX. Typical smoking parameter profiles are illustrated in Figures 6-10. The following are trends which were associated with a majority of the subjects: The more irritating cigarette was associated with: 1) higher flow rates (10/14 7.03% higher) 2) shorter durations 9/14 14.75% higher) 3) shorter Intervals (10/14 15.06% lower) 4) higher TIN ratios (12/14 5.55% higher) (2 subjects did not differentiate between the two cigarettes with respect to irritation). There appears to be some association between Irritation and higher flow rates and Irritation and higher TIN ratios. Since flow rate and T/N ratio were somewhat positively correlated (-f - 0.52), it is most likely that the three are related. It could be speculated that smokers who take more frequent puffs of greater intensity (shorter durations, higher flow rates) may experience greater irritation. Since the ratios of other smoke constituents in the volatile and seml-volatile phases were not measured, irritation cannot be attributed only to the observed trend in TIN ratios. CD %,0 Co Co BATCo document for Legal Services : Health Canada 20 October 1999 There were no particular trends associating objective smoking parameters with easier draw and better mouthfull. This may have been due to the inexperience of the participants with respect to subjective assessment. There was no control over the way in which they may have Interpreted the subjective characteristics explained to them. Draw effort and mouthfull are two terms which may be unclear or easily confused by smokers untrained in subjective assessment. Further work in this area to investigate the relationship between subjective characteristics and objective parameters and smoke constituents, will make use of a panel trained in subjective evaluation. 2. CONCLUSIONS The close relationship between smoking behaviour, cigarette design and subjective assessment has been emphasize n s repor Further study of the exact nature of these relationships, In terms of specific smoking behaviour parameters and subjective characteristics Is required and will be initiated in the coming months. A panel trained in subjective techniques is clearly necessary. It appears that: 1) per puff behavloural changes are a better indication of smoker modification than are total cigarette differences 2) modification of smoking behaviour is used by the smoker to alter cigarette deliveries 3) there is a link between behaviour modification and 18 product preference 4) there is some association between irritation and flow rates and irritation and T/N ratios CD \Z Co "0 BATCo document for Legal Services : Health Canada 20 October 1999 -14- TABLE I DETAILS OF THE CIGARETTES UNDER STUDY MATINtE SPECIAL CIGARETTE FILTER 919 Delivery Data DPM (mg) 9.1 10.1 Nicotine (mg) 0.72 0.88 Water (mg) 1.8 1 .1 Co (mg) 16.2 9.2 Puff Number (mg) 8.5 8.9 Physical Data Filter Type SCS III Filtrona COD Filter Length (mm) 25 25 Pressure Drop: Closed (ins W.G.) 5.0 7.4 Open (ins- W.G.) - 5.2 Ventilation M - 34 Circumference (mm) 24.8 24.8 Length (MM) 84 84 0 CD K) \10 \10 (Z; BATCo document for Legal Services : Health Canada 20 October 1999 00 > q 0 0 CL 0 0 C IIAMLE 11 COMPARISON OF STANDARD MACHINE AND HUMAN MOUTH DELIVERIES Q2 CIGY. AVE RAjL-ALL1XLE"jk1!1!J ---- AVERAGE AVERAGE AVERAGE TAR AND NICOTINE ..... :::: , io ., TAR I I T/N RATIOS Cq/1 RAII95 AVAILABLE PER UNIT CO MC-1 U MN K HUM N MACH 19E JIqMAN MACI JflE_JWAM (D IAR mic JAH NIC HSI` 9.1 11.4(5.4- 3.16)* 0.72 '0.08(0.52-2.01) 16,2 17.2(li.5-34.1) 12.6 12.6(10.3-15.7) 1.8 1.6(1.1-2.1) 0.56 0.04 0.66 0.05 919 10.1 13.4(7.5-31.0) 0.08 1.14(0.61-2.17) 9.2 13.1( 0.2-25.3) 11.4 IT.( 0.9-17.1) 0.9 1.0(0.7-1.2) 1.10 0.101.02 0.09 0) CHIT. AVERAGE & CHANGE RELATIVE TO MACHINE AVERAGE % CHANGE CIGT, 919 RELATIVE TO MAT. $P, F. X TAR NICOTINE CO TAR NICOTINE CO 0 MSF +25 +22 4 6 0) #22 +35 -23 919 -33 #30 +42 06 Delivery range of subjects. 0 0 tMIL 0 Cr (D 66IC4MV - .- -76- TABLE III COMPARISON OF OBJECTIVE SMOKING PARAMETERS FOR HUMAN AND MACHINE SMOKINGS PARAMETER AVERAGE VALUES MATINtE SPECIAL FILTER CIGT. 919 MACHINE HUMAN MACHINE HUMAN PUFF VOLUME (MI) 35 43.9 (21.7-65.9)* 35 47.3 (24.6-68.7) PUFF DURATION (s) 2 1.6 (0.6-2.8) 2 1.7 (0.7-3.7) PUFF VELOCITY (Ml/S 1 , 7.5 28.7 (20.6-38.0) 17.5 29.4 (16.3-39.6) TOTAL VOLUME (Ml) 297.5 473.6 (291.8-1014.3) 311.5 547.6 (339.9-1087.7) PUFF INTERVAL (s) 58 35 (18-69) 58 36 (15-17) PUFF NUMBER 8.5 11 ( 7-19) 8.9 12 ( 6-22) 0 AVERAGE % CHANGE IN SMOKING PARAMETERS .PARAMETER RELATIVE TO MACHINE CIGT. 919 RELATIVE TO MSF MSF 919 PUFF VOLUME +25 +35 8 PUFF DURATION -20 -15 + 6 PUFF VELOCITY +67 +68 + 2 TOTAL VOLUME +59 +76 +16 PUFF INTERVAL -40 -39 +12 PUFF NUMBER +29 +35 + 9 * Delivery range for subjects 40 -Z>. CD r1J N) BATCo document for Legal Services : Health Canada 20 October 1999 -17- FIGURE 1- COMPARISON OF SMOKE DELIVERY RANGES 100% MSF CIGT. 919 56% 44% 31% 25% 25% +10% 1 <-,i 07. 1 > :ti 0% 1 f t 10% 1 <+10% I I *.* +I D% PROPQRTION OF SMOKERS WITHIN THE DELIVERY RANGES C) CD "0 \10 L4 BATCo document for Legal Services: Health Canada .20 October 1999 -Is- _FIGURE 2 STANDARD MACHINE PER PUFF DELIVERY mg/puf f 3.0 2.0 0, 1.0 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 CIGT. MSF PUFF NUMBER CIGT. 919 MSF 919 x NICOTINE r = 0.997 r = 0.99-9 & TAR (PMWNF) r . 0.917 r = 0.977 0 CO r - 0.989 r = 1.000 .9b. CD \-Z \10 Jl- BATCo document for Legal Services : Health Canada 20 October 1999 ca > 1 0 0 CL FIGURE-3 HUMAN PE.R PUFF DELIVERY SUBJECT 4 0 0 mg/puff c 3.0 (D (a co) 2.0 (D CD :r CL 0 Pip, 0 cr 1,2 3.4 5,6 7,8 9,10 11,12 13,14.15 16,17 1.2 3,4 5.6.7 0.9,10 11.12,13 14,15, 16.17 18,19 CIGT. MSF CIGT. 919 PUFF NUMBER x NICOTINE * TAR (PMWNF) 0 Co ~6610peoV -20- FIGURE 4 HUMAN PER PUFF DELIVERY SUBJECT # 7 Mg/ Uff I I 3.0 2.0 0 , 11 1 2,3 4,5 6,7 2,3 435 6j7 CIGT. MSF CIGT. 919 PUFF NUMBER x NICOTINE * TAR (PMWNF) 0 0 CO C3,11% BATCo document for Legal Services : Health Canada 20 October 1999 -21- FIGURE 5 HUMAN PER PUFF DELIVERY SUBJECT # 9 ing/puf f 3.0 2.0 d 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 CIGT. MSF CIGT. 919 PUFF NUMBER x NICOTINE * TAR (PMWNF) 0 CO 0 4>. CD CD -1-4 BATCo document for Legal Services : Health Canada 20 October 1999 -22- TABLE IV COMPARISON OF RESTING LEVEL ALVEOLAR CO VALUE AND RELATIVE INHALATION-FOR EACH SUBJECT SUBJECT ALVEOLAR CO LEVELS (PPM) INHALATION FACTOR (See Appendix) MSF 919 919/328 1 11.0 8.6 1.02 2 47.2 23.2 0.66 3 15.0 7.6 0.54 4 20.4 8.1 0.59 5 28.7 16.2 0.79 6 22.8 7.3 0.45 7 16.6 14.7 0.84 8 27.1 20.1 0.92 9 28.9 14.8 0.72 49 10 29.8 16.0 0.84 11 40.5 -19.3 0.71 12 21.8 19.6 1.08 13 22.1 15.8 1.18 14 15.3 8.5 0.67 15 33.9 21.5 0.73 16 6.2 3.9 0.80 X 24.2 14.1 0.78 AVERAGE (%) 43 DIFFERENCE 0 CD .1.0 -10 CD BATCo document for Legal Services : Health Canada 20 October 1999 -23- TABLE V COMPARISON- OF THE SUBJECTIVE RESPONSES TO MSF AND CIGT. 919 Wilcoxin Rank Sum Test for Paired Data p a 0.05 H 0 - no difference between MSF and Cigt. 919 SUBJECTIVE CHARACTERISTIC T ca1c. T critical AMOUNT OF TASTE 10 21' *DRAW EFFORT 28 26 IRRITATION 15 17 STRENGTH 6 11 SATISFACTION 6 17 MOUTHFULL is 21 'no significant difference. 0 TABLE Va SUMMARY OF SUBJECTIVE RESPONSES TO MSF AND CIG RETTE 919 SUBJECTIVE NUMBER OF SUBJECTS WHO CHARACTERISTICS SCORED MSF HIGHER SCORED 919 HIGHER FOUND NO DIFFERENCE AMOUNT OF TASTE 10 4 2 DRAW EFFORT 6 7 3 IRRITATION 10 4 2 STRENGTH a 3 5 HOUTHFULL 6 7 3 OVERALL PREFERENCE 6 7 2 (FORCED CHOICE) 0 BATCo document for Legal Services : Health Canada 20 October 1999 -24- TABLE VI COMPARISON OF AVERAGE PER PUFF PARAMETER CHANGES STATED PREFERENCE FOR MSF CIGT. 919 MEAN CHANGE* FREQUENCY OF MEAN CHANGE" FREQUENCY OF % CHANGE % CHANGE INTERVAL 14.4 3008 6.6 1203 VELOCITY 8.5 619 4.8 492 VOLUME 22.8 279 9.5 254 DURATION 7.6 127 5.2 213 MEAN CHANGE - GRP. AV. CHANGE 919 RELATIVE T FREQUENCY OF CHANGE - AV. OF IND. VAR. FOR G TABLE VII COMPARISON OF NORMALIZED DELIVERY CHANGES ON A PUFF BY PUFF BASIS STATED PREFERENCE FOR MSF CIGT. 919 MEAN CHANGE* FREQUENCY MEAN CHANGE" FREQUENCY TAR 23.8 545 13.8 104 NICOTINE 26.5 1145 3.5 69 Co -13.8 0.031 -17.5 0.015 10 T/N -2.0 13 -2.8 0 * 919 RELATIVE TO MSF rQ (Z) BATCo document for Legal Services : Health Canada 20 October 1999 -25- FIGURE 6 HU14AN SMOKING PARAMETER PROFILE - AVE. PER PUFF VOLUME PER SECTION MI/puff CIGT. MSF CIGT. 919 110 100 90 80 70 60 50 40 30 20 I il III IV V VI Vil Viii SECTION OF CIGARETTE 10 9 SUBJECT 4 x SUBJECT * 7 o SUBJECT #9 Section #1 Puff# 1.2 Sectlon #1 Puff Section #1 - Puff # I 11 3,4 11 2,3 it 2 111 4,5 3 111 5.6 4 IV 7,8 IV 6,7 TV V 9,10 V 8 V VI 6 Vi 11.12 Vil 7.8 Vii 13,14,15 Vill 9 VII 16,17 BATCo document for Legal Services : Health Canada 20 October 1999 -26- FIGURE 7 HUMAN SMOKING PARAMETER PROFILE - AVERAGE FLOW PER PUFF MI/s/puff 30 X- 20 'x %X 10 CIGT. MSF CIGT. 919 I II III IV V VI VII VIII SECTION OF CIGARETTE e SUBJECT #4 x SUBJECT # 7 o SUB3ECT #9 Section #1 - Puff# 1,2 Section #I - Puff Section #1 - Puff # I 11 3.4 11 2,3 2 111 5,6 111 4,5 111 3 IV 7,8 TV 6.7 IV 4 V 9,10 V 8 V 5 VI 11,12 VI 6 VII 13,14,15 VII 7,8 VII 16,17 VIII 9 z1b BATCo document for Legal Services : Health Canada 20 October 1999 -27- FIGURE 8 HUMAN SMOKING PARAMETER PROFILE - AVERAGE DUARTION PER PUFF S/puf f 4.0 2.0 CIGT. MS-- CIGT. 91S 1 CL 1 11 111 IV V VI Vii VIII SECTION OF CIGARETTE a SUBJECT 444 x SUBJECT * 7 o SUBJECT #9 Section #1 - Puff# 1,2 Section #1 - Puff Section #I - Puf f I 1 3,4 11 2,3 11 2 111 5,6 111 4,5 111 IV 7,8 IV 6,7 IV V 9,10 V a V VI 11.12 VI VII 13,14,15 VII T's VII 16.17 VIII 2 -P-1. C-D BATCo document % for Legal Services : Health Canada 20 October 1999 -28- FIGURE 9 H_UMANsi40KlNG PARAMETER PROFILE - AVERAGE INTERVA S /puf f CIST. P CIG-1 - 9 'A, -~, 60 40 30 20 10 I il Ili IV V V 1 VU Vili SECTION OF CIGARETTE # SUBJECT #4 x SUBJECT # 7 o SUBJECT #9 Section #1 - Puff# 1.2 Section #1 - Puff Section #1 - Puff # 1 11 3,4 11 2,3 11 2 111 5,6 111 4,5 111 3 IV 7,8 IV 6,7 IV 4 V 9,10 V 8 V 5 V 1 11,12 VI 6 VII 13,14,15 Vii 7,8 V11 16,17 Vill 9 4~--b CD Kj BATCo document for Legal Services : Health Canada 20 October 1999 -29- TABLE VIII COMPARISON OF THE VARIANCES OF THE PER 2UFF DELIVERIES AND DELIVERY RATIOS VARIANCE OF THE PER PUFF VARIAKE OF STATED TAR DELIVERY NIC. DELIVERY T/N RXTIO PREFERENCE CIGT. FOR SUBJ4 MSF 919 MSF 919 MSF 919 1 0.4504 0.4725 0.0570 0.0510 2.1767 4.7367 MSF 2 1.1241 0.8038 0.0557 0.0832 5.1000 6.1857 14SF 3 0.5892 0.6584 0.0510 0.4321 1.6267 0.4958 919 4 1.1611 0.38S8 0.0410 0.0739 2.0314 10.4141 MSF 5 0.2439 0.2699 0.0311 0.0320 3.3092 3.6100 MSF 6 0.7373 0.3026 0.0480 0.0251 8.0967 0.6867 MSF 7 0.5115 0.3202 0.0436 0.0386 0.9292 5.2700 919 8 0.2360 0.3399 0.0224 0.0151 8.5936 1.2580 919 9' 0.4657 1.0802 0.0389 0.0639 8.2002 4.7394 919 10 0.5367 0.5228 0.0597 0.0408 4.400 3.333 MSF 11 0.6521 0.9581 0.0504 0.0727 2.2995 1.9670 None 12 0.8091 0.6706 0.0605 0.0551 0.4625 2.7500 MSF 13 0.4500 0.7131 0.0420 0.0560 0.9166 6.2500 919 14 0.4876 0.4741 0.0346 0.0403 2.1692 1.7492 None 15 0.5417 1.3106 0.0632 0.0853 8.6490 3.0698 919 16 0.1761 0.4436 0.0206 0.0411 7.2833 1.6492 919 AVERAGE VARIANCE FOR PER PUFF STATED PREFERENCE FOR MSF OTHER(919) 919 OTHER(MSF) TAR DELIVERY 0.7209 0.5242 0.6461 0.4634 NIC. DELIVERY 0.0502 0.0560 0.0947 0.0412 T/N RATIO 3.013 6.796 3.183 6.374 C) C~ t.n BATCO document for Legal Services : Health Canada 20 October 1999 -30- FIGURE 10 AVERAGE HUMAN T:N RATIO PER PUFF CIGT. MSF CIGT. 919 30 20 10 -,e 1 11 Ili IV V VI Vil Viii SECTION OF CIGARETTE * SUBJECT #4 x SUBJECT # 7 o SUBJECT #9 Section #I - Puff# 1,2 Section #1 - Puff # I Section #1 - Puff # I 11 3,4 11 2.3 11 2 111 5.6 111 4,5 111 3 IV 7,8 IV 6.7 IV 4 V 9,10 V 8 V 5 VI 11,12 Vi 6 Vil 13.14,15 Vii 7.8 Vil 17,18 Vill 9 .C:b. C.D CD N) C:D CD C> BATCo document for Legal Services : Health Canada 20 October 1999 -31- TABLE IX COMPARISON OF CHANGES IN OBJECTIVE SMOKING PARAMETERS AND SUBJECTIVE RESPONSES I MEAN CHANGE IN I CIGARETTE THAT WAS/HAS ~UBJECT VOLUME VELOCITY DURATION INTERVAL T:N I MORE EASIER BETTER - 1IRRITATING DRAW MOUTHFULL 1 27.9 18.05 7.13 -16.73 22.69 NSF MSF 2 -3.55 2.08 0.30 53.81 -3.43 919 919 MSF 3 -8.78 4.74 3.08 -0.68 2.99 919 919 919 4 -4.96 1.43 -6.98 -13.18 9.61 919 MSF MSF 5 14.88 -4.80 11.91 -6.81 -7.95 MSF MSF MSF 6 -10.97 -13.01 5.54 -27.24 -11.72 919 919 7 -4.36 -5.95 -11.15 -8.17 -3.64 MSF MSF 8 -15.56 -11.83 9.58 -31.36 4.61 MSF 919 9 36.44 -6.19 47.88 9.65 -3.85 MSF 919 919 10 2.59 -9.71 14.67 -27.57 -2.66 MSF ASF 11 1 40.02 -10.54 18.68 20.17 -9.80 1 MSF 2 MSF 1 1 12 -2.48 -13.06 -1.94 27.80 -4.68 MSF 919 919 13 25.32 6.61 13.16 -8.86 -1.43 14S F NSF 919 14 14.38 21.03 9.13 -4.45 7.73 919 919 15 33.44 25.23 -1.78 20.04 -1.66 MSF 919 MSF 16 15.51 11.44 2.50 12.19 -6.58 MSF 979 #… CD C:D I\j BATCo document for Legal Services : Health Canada 20 October 1999 -32- APPENDIX I METHODS 1.1 Clqarette Selection The cigarettes used In this study were MATINEE SPECIAL FtLTER (the control), which is a non-ventilated brand, and a sample coded 919, which used a ventilated, high pressure drop filter. The two cigarettes were identical in construction except for the filter; details are given in Table I. The cigarettes were weight (ñ 15 mg) and pressure drop (+ 0.5 in. w. g.) selected. 1.2 SubJect Selection Subjects who participated in this study were selected using the following criteria: 1) their own brand was Matin6e Special Filter or a non-ventilated brand in the 8-11 mg tar range. 2) they smoked at least 10 cigarettes/day. 1.3 Monitorinq In all, sixteen smokers participated in the study (13 female, 3 male). All subjects were allowed unlimited access to the cigarettes. After a weekend acclimitization period, the smoking patterns of each subject were recorded 3 times/day for five days with each cigarette. Samples of alveolar CO were collected, using the re-breathing technique, prior to each cigarette of the day (after 10-30 A.M.). All butts were collected, measured for butt length and analysed for nicotine content. During the second testing session the subjects were asked to complete a questionnaire on their subjective impression (see Appendix II). 1.4 Duplication For each subject and for each cigarette, fifteen smoking profiles were recorded. Using the program AVG on the smoke duplicator, which allows the user to average smoking patterns across brands and/or subjects, a single Oaverage" profile was obtained for each subject with each cigarette. The average profiles were duplicated an a total cigarette and on a per puff basis. The I.T.L. Smoke Duplicator was adapted to reproduce on a per puff basis by the addition of "on-off" toggle switches to each of the motors of the four ports. In this way, each motor could be activated selectively to smoke specifically designated puffs. Each profile was reviewed for puff duration; interval and number of puffs. Where the puff duration was less than 0.7 seconds, the puff interval less than 10 seconds or the total number of puffs was greater than 16, two or three puffs were combined. The first and last puffs were always duplicated as single puffs. Tar, nicotine, water and CO deliveries were determined for both methods of duplication, using standard analytical procedures. _C~:- CD C:D cz~ BATCo document for Legal Services : Health Canada 20 October 1999 -33- 1.5 Data Analysis 1.5.1 Profiles The behaviour and deliveries associated with the recorded profiles were compared: 1) with the standard machine conditions and deliveries on a total cigarette basis (Tables II & 111). 2) to each other for both cigarettes and all subjects on a total cigarette basis (Tables II, III & Figure 1). 3) to standard machine conditions and deliveries on a puff by puff basis (Figures 2,3.4 and 5). 4) to each other for the two preference groups on a puff by puff basis (Tables VI and VII). I.S.2 Carbon Monoxide The collected alveolar CO samples were analysed using a BENDIX CO analyser. The average levels for each subject and cigarette were determined and compared to each other using a Relative Inhalation Factor (Table IV). It was assumed that the amount of expired alveolar CO was proportional to the amount inhaled. From this assumption, COa ct kACQm where COa - alveolar CO; C01n = mouth CO; A is a constant specific to an individual's metabolism. environment etc. and k is the fraction of the total volume of smoke in the mouth that is inhaled. For each cigarette and each subject, therefore: COa(MSF) a kACOM(MSF) and COa(919) a kWC0m(qlg) ~Oa(MSF) M kA and CDm(MSF) COa(919) k'Al com(919) Both kA and VAI are indications of the degree to which smokers Inhale, therefore the degree to which a subject inhales one cigarette relative to another can be indicated by the following: COa(MSF) COM(MSF) kA = Relative Inhalation Factor COa(919) k'A' com(919) Ni 4-1>. CD N) C-- CD \-c- BATCo document for Legal Services : Health Canada 20 October 1999 -34- The average duplicated CO deliveries for each cigarette were compared to the standard machine deliveries and to each other (Tables 11 and III). 1.5.3 Sublective-Impressions The responses from the subjective impression questionnaire were analysed using the Wilcoxin Rank Sum Test for paired data (Table V). Based on the responses, the smokers were divided into two preference groups (Table Va). CD CD CD BATCo document for Legal Services : Health Canada 20 October 1999 -35- APPENDIX II SUBJECTIVE QUETTIONNAIRE Please evaluate the two cigarettes you are smoking in this study on the qualities listed below by circling a number on each line which best describes how you rate each cigarette. A LOT OF TASTE 7 6 5 4 3 2 1 NO TASTE IRRITATING 7 6 5 4 3 2 1 NOT IRRITATING EASY TO DRAW 7 6 5 4 3 2 1 AIFFICULT TO DRAW STRONG 7 6 5 4 3 2 1 WEAK VERY SATISFYING 7 6 5 4 3 2 1 NOT AT ALL SATISFYING A LOT OF SMOKE IN 7 6 5 4 3 2 1 NO SMOKE IN MOUTH MOUTH Please choose one of the two cigarettes as your preferred cigarette MSF 919 What are your main reasons for this choice? it -r-1~!- CD CD BATCo document for Legal Services : Health Canada 20 October 1999 -36- APPENDIX APPENDIX III CALCULATIONS FOR MEAN-PARAMETER CHARGES AND FREQUENCY OF CHANGE TERMS The following is an example of the calculations performed on each of the sixteen subjects participating in this study. Each relative change in the parameters was calculated as: the parameter value associated with cigt. 919 the parameter value associated with ci9t. MSF- Subject #9 1 % Change in Parameters 919 Relative to MSF Puff * I Volume Mean Flow Duration Interval I (MI) - (ml/s) (s) (s) 9.69 20.00 39.47 17.07 2 9.83 22.64 33.33 -29.10 3 18.92 9.09 30.00 -24.59 4 10.07 19.05 31.03 -29.23 5 1 21.14 11.68 37.04 41.30 6 54.08 3.45 60.87 59.46 7 82.86 -13.76 57.14 80.00 8 62.50 15.34 94.12 X 30.97 10.94 47.88 16.42 Var. 781.9799 139.1809 485.167S 2058.6833 X "Mean Change" ( I Variance - "Frequency" in Tables VI and VII Following this method the "Mean Change' and "Frequency" were calculated for each subject. From the subjective responses the subjects were assigned to one of the two preference groups. Each group's average value for both "Mean Change" and "Frequency' was then calculated for all the puff parameters. the same method was used to analyse the per puff human deliveries for the two preference groups (Table VII). _r_ CD rQ BATCo document for Legal Services : Health Canada 20 October 1999 CD Q-4 BATCo document for Legal Services : Health Canada 20 October 1999