Tobacco Documents Online

~en:ephaingram in srnuking ( 20. Smoking behaviou, in Germany-the analysis of cigarette butts (KIPA) •.v SCwUt.ZAVD F SEEHOFEA Reproduced from Smoking Behavior: Physiological and Ps cholo icalInfluences (R.E. Thornton, ed.) E inburgh, Churchi -Livingstone, 1978) Smoking beiuviour in Getmatty 1968-1977 When we started' our work concerned with smoking behaviour, in 1968. we had ventured into new territory. The only known factors were some results from observations and puff profile recordings of smokers (Keith and Hackney. 1962) which may have influenced machine smoking conditions. At about the same time measurements and observations were made to detertnine differences in smoke deliveries between'free smoking' or natural smoking, and 'restricted smoking' or machine smoking. Additionally ideas were put forward to estimate smoke intake by analysing the amount of smoke retained in the tobacco butt and on the filter tip. There were also first thoughts and proposals for the reproduction of human smoking by means of a'slave•smoker' or'puff dupiicator',. Our ideas stemmed from'the observation that the range of human smoking behaviour is very wide and we would have to study many subjects. Experimental methods should be chosert or developed which would allow us to draw conclusions about the behaviour of the average smoker. Of all methods available, we thought that the examination of cigarette butts disauded by smokers would be most successful. This was the only totally non- tnvasive way to estimate the amounts of smoke taken by human smokers, and such cigarette butts are available in large enough samples to answer many questions. The butt analysis 'KIPA' (from Kippen, a slang term for cigarette end) is limited to filter L-igarettes, which formed approximately 35 0 of the cigarette consumption in Germany. The basis of KIPA is the correlation of the amount of smoke retained in the filter with the actual amount of smoke taken in by the smoker. Under constant smoking conditions (puff volume and puff duration, which is equivalent to constant tlow rate) the ratio between the amount of smoke retained in the filter and the actual intake of smoke is constant. One can therefore calculate the amount of smoke taken by the smoker from the amount of smoke retained by the filter. Unfortunateiy. this is not the case in natural smoking. Smoking conditions arc not constant, and the ratio of the smoke retained by the tilter and the amount of smoke taken by the smoker is not necessarily constant. Additional information about smoking behaViour and knowledge about the Jependance of filtration .oefficients t for indivuiuai smoke cumponentsP on smoking :onditions are therefm necessary. Only then can the actual intake of smoke be •cu

:60 &MOK1NG BEHAVIOUR caIt.~ttlated. .4 simple approach would be to.examine the ratios of various smoke components in the nlter (e.g. condensate, nicotine and phenols) with respect to puff volume, puff duration, puff frequency and butt length(for single cigarettes) and to check that these ratios remain constant for cigarettes smoked naturally. By use of the appropriate retention value the actual amount of the smoke component taken :ouldbe calculated. The values obtained by this method would be hypothetical and would need to be checked by machine smoking, using the calculated conditions. It is also possible to assume initially that some smoking parameters, e.g. puff duration. puff interval, puff number and butt length, are fixed values which may be obtained from sufficient observations. An average puff volume can be calculated from a sutficient number of flow profile recordings which would be exact enough to check the assumed values. The first tests in the years 1968-71 showed that such simplified test models are not sufficiently accurate to describe natural smoking. The lack of consistency in smoking conditions whilst the cigarette was smoked proved to be the biggest problem. Butt length In butt analysis it is assumed that there is a relationship between the amount of tobacco smoked and the amount of smoke retained' in the Mter tip. It is known from manhine smoking that per puff smoke dellveries inaasr as the dprette is smoked. it:an be assumed that the last few puffs are very Important for the smoker and that small differences in butt length can have considerable effects on smoke deliveries. How do these presumptions agree with reality? Butt lengths of ?2 nun for filter cigarettes and 19 mm for cigarettes without ftlten. were found in West Germany by von Bechrnann (1959). Ten years later we noted a butt length of 30.6 mm for filter cigarettes and 25.6 mm for cigarettes without fiiters ISchulz and Seehofer. 19701, although average smoke deliveries, especially in hiter cigarettes, were lowered dramatically in the meantime. Butt lengths increased until 197: (Schuiz. 1974) but have since decreased, probably for economic reasons such as an increase in tax and a recession (Schulz. 1974; Schulz and Seehofer. l?; 6). T.able :U.I Butt lengths in West Germany Filter Cigarettms Cigarettes arithuut filter Number Butt Length Number Butt Length umpled' ImmY tampted (mm) )3~ 1! __., 19.3 i9Sa; ?q.6 939 25.4 , e. These tests show that the rang, considerable. However, it was p, large numbers of butts were coil- length for different sections of t. location, and economic conditlo 1972; Schulz and Seehofer, 197 these differences with respect to An analysis of the amount of : can give very variable results for in..rease during the last few mill. was too low to be detected duri reason it is necessary to proceec different smoking classes. The shows a uniform picture • the ir, srnall during the last few puffs. last part of the cigarette and thc (Figures 20.1 and 20.2). Observed smoking parameters PutT number, puff interval and by observing smokers with timi constant run speed etc.). It is c as the actual drawing of the pu: cigarette is held in the mouth o be determined by observation. In two experiments csrried o smokers l two groups of 100 an number, puff duration and puf as the time for which the :-4gul surreptitious tests it was :mpos Observations were made in put which means in nearly all situa for puff number. puff 3L!:itior of !hese values show a remark;: long tune period. especially th• .alues si:•.)wn in Table .U._. s T.tbie : Q.: SmoKtn; ; ~:a~e:a a.stf N'umbt: ...~ ~Q~~ ~.).. -tln:'i.~ N ~lt . 1.. . ... ~. ~

.. • .:S.k l'r SlS t)F C'K:A K1:TTE liilTfa ar~ . • .. ~ - s ot'various smoke components witkx-spect to puff volume, ;ing' irettes) and to chedc ~ ted `ta.unlly. By use of the e smoke component taken :thod would be hypothetical . using the calculated conditions. := cing parataeters, e.g. puff dtuation, • f~P be obtaiaed ~' i values whiclt ma y , ,te can benlcvlated from a would be exact enough to check • 8•71 showed that such simplified : natural smoking. The ladc of tte was smoked proved to be :ship between the amount of ;' in the filter tip. It is known ies inaease as the cigarette is are very Important for the , .t have considerable effects on e with reality? t tttm for cigarettes without ann (1959). Ten years later ttes and 25.6 mm for cigarettes igh average smoke deliveries, ,.:11I he meantime. Butt ve st- decreased, probably for ecession (Schulz, 1974; Schulz Cigarettes without filter Number sampled Butt Leneth (mm) 19.3 935 25.6 4364 29.2 3623 28.0 7099 26.3 fhese tests show that the range for single but. engths for different brandi was considerable. However. it was possible to :alculate very exact average values as large numbers of butts were collected. There are significant differences in butt length for different sections of the population, e.g. sex, place, occupation. geographic loeation, and economic conditions. brand etc. (Schulz and Seehofer, 1970; Schulz, 1972, Schulz and Seehofer, 1976). The question arises as to the importance of these differences with respect to the amounts of smoke drawn. An analysis of the amount of smoke retained in filters with respect to butt lengths can give very variable results for some smokers, ranging from a disproportionate increase during the last few millimeters of the cigarette smoked, to a level which was too low to be detected during the last third of the cigarette smoked For this reason it is necessary to proceed with different types of smokers typical of the different smoking classes. The average of all smokers or large groups of smokers shows a uniform picture • the increase of smoke retained in the filter is only very small during the last few puffs. The average smoker puffs only a little during the last part of the cigarette and the amount of smoke taken from these putTs is small (Figures 20.1 and'_0Z). Observed smoking patanteters Puff number, puff interval and total time alight can be measured relatively easily by observing smokers with timing devices (stopclocits, cassette recorders with constant run speed etc.). It is considerably more difficult to measure puff duration as the actual drawing of the puff does not have to be dential with the tuae the cigarette is held in the mouth or with a visible onset of glow. Puff volume cannot be determined by observation. In two experiments carried out in 1971 and 1974 in Hamburg a large number of smokers (two groups of 100 and 218) were observed surreptitiously and puff ntlmber, puff duration and puff interval were measured. Puff duration was defined as the time for which the cigarette was held in the mottth. In most of these stureptitious tests it was impossible to determine the brand name or butt length. Observations were made in public houses, railway stations, on the road or at work, which means in nearly all situations possible. Table 20.2 shows the average values for puff number, puff duration, puff interval and total of the puff durations. Some of these values show a remarkably high consistency in smoking behaviour over a long time period, especially the value for total puff duration which, of all the values shown in Table 20.2. is most relevant to smoke deliveries. Table 20.2 Smoking parameters of smokers observed in Hamburg r911 and 1974 Puff Number Puff Duration (see) Puff Interval (sec) Total Puff Duration (sec). U971 1974 1971 1974 L971 1974 1911 1974 Nen 10.: 10.9 1.47 1.47 52.9 42.1 ' 13.0 16.0 Women 10.9 13.3' 1.31 1.17 46.0 40.7 14.3 15.5 All 10.5 11.8 1.41 L34 50.3 41.5 1e.8 15.9

262 $AIOKING BEHAVIOUR AMOUNT OF NICOTINE RETAINED IN THE FILTER WITH RESPECT TO BUTT LENGTH NICOTiNE ( m9/TiP) i •oo t 0• 90 0•80 0•70 0.60 0.50 0•40 Y x AVERAGE VALUES FOR ALL. ~ 5UB,IECTS (SERIES I) 0 5TANDARD MACHINE SMOKING CANOITIONS (FRM 33R1s.IN 2•a st.4 I PUFF/ min) ~ MAGiINE SMOKED CIGARETTES . .. E. .~ UNDER CONDIT10N5 OETERMINED FOR 9U6JECT5. (fRFE, 50mEs IN I• 6 stt. 2PUFFS/min) i-• { - ~ .. ~ ZO 25 30 35 40 3UTT t-ENGTH (tr+nyl AMOUNT OF TPM W(TH RESPECT TPM ('ng f TI P) A;b

.%s.36 LtblScii %'i!. .4R; t'''_ irL fTS :c: IED IN THE FILTER T LENGTH ~ VALUES FOR ALL 3 (SERIES I) t0 MACHINE 'SMOKING NS (FKE~ 35nts W Z-0 xc, : SMOKED ClCiI1RETTES ONDITIONS OETERMlNED 1 ECTS. (FREE, 50 n+!'s IN '-:PUFFS/m1n) . . AMOUNT OF TPM RETA(NED IN THE FILTER WITH RE5PECT To BUTT LENGTH x AVERAGE VALUES FOR ALL SU6JECT5 (SERIES I) 0 STANDARD MACHINE SMOKWG CONOiT10N5 (FREE,35rals IN 2-0stc I PUFF/mun) p MACHINE SMOKED CiGARETTES UNDER CONDITIONS DETERMINED FOR SUBJECTS. (FREE, 50mes IN I• 8 SCG, 2 PUFFS/ min) TPM (rn91TIP) 151 1o 5 3s 40 20 25 30 ~(mq _ 6U TT LEN GTH (m m) Fi;. 20.2

I.. 264 SMO1C1NG BEIiAViOUR There are some significant differences in observed smoking behaviour between men and women. Women leave a longer butt (on average approximately 2 nttn longer) and have a shorter puff duration than men; they seem to compensate for this with a higher number of puffs and have therefore the same total puff duration (puff " duration x puff number). There seems to be a trend for a shortening of puff dtuationn with increasing age in tnen. In women however there is no such difference in smoking behaviour (Table 203). Table 20.3 Smoking parameters with respect to a9e (1974) (average number of all subj.cts 16-36) Puff Number Puff Duratioa Puff Interval A e (sec) (Kc) s Men Women Men Women M+n Women 15-25 10.7 13.2 1.57 1.17 39.1 39.5 25-35 11.2 14.6 1.49 1.15 39.6 3.9.6 3 3-4 S 10.7 13.6 1.42 1.21 42.8 40.3 45-55 10.7 13.0 1.40 1.16 41.3 38.6 SS-7S 11.4 - 1.37 - 48.0 - No differences were observed between cigarette brands or different types of cigarettes as far as smoking behaviour was concerned with the exception that puff durations were clearly longer when cigarettes manufactured from air-cured tobaccos were smoked (1.8 instead of 1.4 sec). It was observed that smoking in situations that caused' physical stress (e.g. when waiting) resulting in a shortening of puff interval' (35 instead of 43 sec). In men the puff duration was shortened as well (puff duration 1.25 instead of 1.38 sec). All the values discussed above are averages of how a whole cigarette was smoked. The distribution of the values for single puffs (puff duration and puff interval) shows a definite trend with respect to puff number (Figures =0.3 --0.5). Men and women show z trend for a shorter puff duration with increasing puff number (Figure ?0:3). The frequency with which :ong duration puffs occur decreases from the first to the !ast:ufft the frcauencv :i •x:.urrence ot' Ehe shortest puff is :auallv low durinn the •'!rst ive auas '>uc nr: ase•s r-rn then,m until the !ast puff ('Figure :0.4). Puif interval increases •.v;th :n:re:utng Puit ~:ur.::,er ! -iYUre :0.:1. :n aummar,. one ;,-an deduce from s..ec:eastng puff litratwr. !nu in :n,;reasing putT interval s 3ecreasing smoking :ntensit;: 3ependent on purf:uraber. T•nese results agree wit.'t :ae small increase -9C smo1C' 'etatned in the !ilrt: ~ur!Gi :Tlt :~it it%v puffs. Smoking parariieters from puff profile recordings `•Vlttn outf arotiles ire re-:..rsec .vitE: a!it'crenriat cr.isur•_ !._~rtn_ .nd Searaier. :?-: .!te .-tecsi: • :, "Jrfs 1nU rhC -!lSrflbilrr ' )1*:he incenSit'• 1"arif'gi ... .ulatiJft: :r:d 7ur': PUFF DURATION Wl PUFF DURATION (sec) I-b+ 1-4t 1•3 t I•I t I•0 I 2 3 PU fig. 20.3 intervais. Important data can " The puff volume may 4e .._-cais rate can be obtained from the •. +olume •xas drawn can also be 3uc:~ profiles xere recorded ;97?. The values abtaine1 fro :ations made but Sisag ee xir:: :5 to 19 puffs were recorar: anoicersf when puff protik :ec -ecordina situationss or •.vrlea ... -Prcceiure an:i the -ec,,reing .: whtcf: an be :uite ::snstder:i- :~)-3: :a.. a ,:Ompacibi. xr,i: : Gst`anJtd .,auCS -7CtaIneQ '• .f •• 20215'74s5s

ed;smokiag-behaviour between men ige dpproxiautely 2 mm longer) s~~ c~ompensate for this with :am`' ~ I puff duration (puff v:nd ~'or 1 shortening of puff duration tere is no such difference in xrtoking ge (1974) i Puff Interval (sac) ietr Men Women 39.3 39.3 39.6 39.6 42.3 40.3 41.3 38.6 44.0 ~ brands oc diPferent types of ned with the exception that puff tufaetured from air-cured tobaccos caused physical stress (e.g. when (35 instead of 43 sec). In men the ~ n 1 Vead of 1.38 sec). )w a ole cigarette was smoked. ff duration and puff interval) shows ires 20.3 • 20.5). f duration with increasing puff t long duration puffs occur decreases ~:currence of the shortest puff is ss from then on until the last ber (Figure 20.5). In summary nd an increasing puff iruerval a iumber. These results agree with luring the last few puffs. I pressure transducer I e.g. Lorenz .ntensity of puffs and the distribution umbers, puff durations and purf .~,. ANALYSIS OF CIGARETTE BUTTS 265 PUFF DURATION WITH RESF. ,T TO PUFF NUMBER PUFF DURATION (sec) I•6+ 1'4 t 1.3 t 1•2 t I•0 Fip 20.3 o MALE x FEMALE 2 3 4 5 6 7 8 9 10 11 12 . PUFF NUM6ER interv:ls. Important data can be deduced from a knowledge of the puff profile. The puff volume may be obtained by integration of the profile uea, the peak flow rate can be obtained from the height of the profiles and the time in which the volume was drawn can also be obtained. Such profiks were recorded from a large number of stttokers in 1970, 1971 and 1977. The values obtained from these experiments agree with•some of the obser- rations nude but disagree with others (Table 20.4). 15 to 19 puffs were recorded as average values for larger groups (about fifty smokers) when puff profile recordittgs were made compared with 11 puffs in other recording situations or when observed surreptitiously. This shows that the survey procedure and the recording situation have an influence on smoking behaviour which can be quite considerable. The puff intervat, which reduced from 42-50 sec to 30-35 sec. is compatible with the increased number of puffs. lt is difficult to estimate values obtained for puff volume as comparative values could not be obtained from surreptitious observation. Three experiments in 1970-77 showed between

266 SMOKING BEHAVIOUR FREQUENCIES OF THE LONGEST(A) & ~- SHORTEST (8) PUFFS MALE AND FEMALE FREQUENCY (0/0) i PUFF INTERVAL NUMBER OF INT PUFF INTERVAL (sec) 4 5 n-4 Ir3 n-Z n•1 n Sot ~ PUFF NUMBER FiB. 20.4 Tabl. 20.4 Smoking parameters from observations and profile recordings (Meaa values for groups of 30-218 subjects) r:: ~. Observations ProfW Recordings Parameter . 1971 1974 1970 1971 1977 Pu ft nu m b sr 10.5 11.8 1549 9 l 4 11 Puff duration 4aecl 1.41 1.34 1.3 1.7 - ~ Puff interval (sec) 50.3 ;1.5 30 3S - Putf rolume (ml) - - SO 60 55 50 and 60 ml per puff l7able .0.4). Most smokers seem to reduce their puff mlume after the !itst three to four.puffs. This reduction could not be ;ound n 3ur profile recordings whick su¢gests that the reasurement ind recording procedure .ntluences not )niv vuff number but =ls-i vuiif luration ind our` •:olsrne. .. 30 vo JF

ANAL) 5I5 OF CIGARETTE BUTTS 267' EST_I-4) & ~ W~ FEMALE ,_ _- Pt1FF INTERVAL WITH RESPECT TO THE NUMBER OF INTERVALS BETWEEN PUFFS ~t---~c FEMALE o---o MALE PUFF INTERVAL (stc) 1-3 n-z n-i n ~ so+ ;d protile r.cotdiap e R.eordiA~s 1971 1977 14 11 :.: 6 ~ 1.7 - .-y. 33 - 60 55 at to reduce their puff ~ tion could not be found in 30 .~' xnent and recording procedure )n and. puff voiume. ~ F1g.20.5 ~ :C - J . Z 3 4 5 6 7 8 9 10 II No. OF PUFF INTERVALS

Smoking parameters, as deduced from the amount of smoke retained in the butt Assumptions about the smoking behaviour, especially the average puff volume,, cout be made from the amount of smoke retained in the butt. It is, however,, necessary to know, from comparable tests, the dependence of the amount retained in the cigarette butt at various flow rates for several smoke components. For this reason we determined this relationship for nicotine, condensate and phenol retaiited' in the butt of one test cigarette (German blend with usual cellulose acetate t'dter). It can be seen in Figura 20.6 - 20.8 that for this cigarette the amount of phenol and condensate retained, with respect to puff volume (constant puff duration and frequency) was similar, apart from a slightly different gradient. With decreasing butt length and increasing puff volume the amount of phenol and condensate retained increases disproportionately. This disproportionate increase was not seen in the amount of nicotine retained. It`was examined under conditions similar to human smoking and the ratios of nicotine: condensate: phenol found were 0.5:12:0.13 The results showed that these values coincide with a puff volume of more than 65 ml and a butt length of at least 43 mm (Figures 20.6 • 20.8). The smokers in this test group must have taken a very large volume of, on average, approximately 65-75 ml during the first few puffs. During the last few puffs (at least from a butt length of 43 mm on.vards) they could have hardly dra.vn any unokeat aIL The true average volume for the whole cigarette smoked in these tests does not seem to have greatly exceeded the standard volume of 35 ml. To confirrrt this hypothesis, special smoking engines (puff dupliator, slave smoker and others) are necessary which are able to smoke to such programs. Amounts of imoke drawn from the cigarette Retention coeffIcients were ob tained for the filter tips using the smoking parameters already mentioned. The amount of actual smoke taken from one test cigarette (Figures ?0.6 - 20.8) was then calculated (Tables 20.5 (a) and 20.5 (b) ). Tabie 20.5 (a) .4verage smoking parameters of test smokers (see Table 20.4, 1970). Puff duration (see) 1.8 Puff interral (see) 30 Puff volume (tnl). S0 Butt lenfth (mm) 31 Table 20.5 ib! kctuai'amuunt at' imoYe taxen by test smokers (from KIPA vaiues) (t:0 subjecu. imoke :omoonenu Amount ot smoke Retention Coeffieienu Actual amoun; retained :n :he butt t:om Table :0.J !a) tmuke intake iX:PA +aluest img) + -.mVet;) Yiratine 11.51 40 T~)t]1 7nln~?li !.. . ij AMOUNT OF NICOTir WiTH RESPECT TO 6l NICOTiNE (m9/TIP) 1•0 0•5I N O N r ~ ~ ~ O