Tobacco Documents Online

I I I I I I I I I I I I I I 1 I eruuA fwn..l.f Al1ua.. 77 (1982) 1145-158 © 1982 Society for the study of Addiction m Alwhol and other Dru;s. 0007 4A90/d2/QO 19-0145 f02.00 Long-Term Switching to Loyv-Tar Low-Nicoti.ne Cigarettes M. A. H. Russell, F.R.C.P., F.R.C.Psych, Senior L.ecturer, S. R. Sutton, Ph.D., Psychologist, R. Iyer, Statistical Assistant, C. Feyerabend BSc., Senior Biochemist', C. J. Vesey M.Sc., M.L BioL, Research Assistaat' Addutioa Rssearc/t Uuit, lustitute of Psycliiatry, Losdos SES 8AF Summary Pxjfis j,%tsers.ad bl.od>wasrra.f ssob istarte tnore stsdied is 1Y serbj+cts kf.n sd.f ter smitcA=q &loan-ta (10.9 xj) lom-sicotise (0.7stj) cilaettes f.r 12 sve+ats. Puff rste, /WffaeGose, totd aolsaru Nffed jer erj jaette iad tlae sssikr of ci jntttes ssubed mm sN ui.ued si,pajitsstlj 4fter smitc*j to t+4 loawyield cqaettes. T<ae ndreti.. iie srfttA-lsarl rsiatiae iataEe (eslcrlaed fi..e brtt cattest) was Nkxtical So tla 47 jer cest eedsctioa is the wschiar-ssaked sisetiaerXds. Ms ic in teejest ivith the sstmtrsu of ~/'ssj is shmW tir .bseacs of sy asdeacj a nVeles.k for the rrdwad yirllr h wcntsis; dr iRte" of Pwffiq. Ir: caitrast, plasrna aicotiuu aad c.titeixe /eaels asre redrud proportiosally kss tkox the soaemal yieldr isdicstis( that Ae srbjects cowepeisseted to sm eztest by wreasit;l neltalttiois. Tlic saersje rtttrctinet ia jlstws rticotis~r ad caisise Were idestiesl (Japa aat). Plasnsa tkiocya.ate sid CONb leatls mere sot sipifuastly deretit sftsr smitckiu to the loanerpield ei jaeettu but the estiuRalsd tas' iatahc to the lwc js &W reduced bJ about 13 jer ctst. A1tbsqi * /wcyield aidvettes run r.ted as slqArtly 1ae sAeit, Ae+r aw rse sipafueat lss .f otti.tf.cti.s. T4 fict tAst arif etsr s.lerated a J19 per cent nd.ctis a #ans sicsti.r roithst ltfst of sairf.cti...nd eoitiwt pq(fnq awr aasessisey s>ICUests tbt the isrnesss ia i.i.lstm aw sot &amad y t ssed ta arataie 4rieatisr rstab. ; This ssd the discrepauies ktmetu aatrssl s.d erperixtental smitcllistg shrdies ae discrssed. Iatroduction It is widely believed that a change to low-tar and low-nicotine cigarettes reduces the health hazards of cigarette smoking. It is not known, however, to what extent the benefits ofsuch a change are offset by the tendency for smokers to compensate for the reduction in tar and nicotine yields by smoking more cigarettes or by increasing the intensity of puffing and inhalation. At present the evidence is somewhat conflicting [l-3). Two different approaches have been used to study this question. One approach has been to compare blood nicotine and carboxyhaemoglobin (COHb) levels of smokers who regularly smoke low-tar low-nicotine cigarettes with the levels of those who smoke brands with higher yields. Two such studies using large samples have found very little diBerence in the blood nicotine and COHb levels of the two groups indicating that their smoke intake was similar despite large di&rences in the yields of their cigarettes [4, 51. It appears, thereiore, that those smokers who have '.*bimm Ueic, New CIat Ho.petal, Leedoa SEi4 ' Asreuhetic >lesumb Laboesoo.y, si. s.rtLoiomew9 Mospiral, r..de. Ee,. C I

I I I I I I I I I I I I I I I I 146 Rursell, Sietton, I~, Fryerabtrid -and Vesey themselves chosen to smoke a low-tar low-nicotine brand ('natural switching') alter their smoking pattern and in so doing compensate almost completely for the reduced yields of their cigarettes. The other approach used to study compensation has been by experimental switching. With this method smokers are required by the conditions of the study to switch from their usual brand. In contrast to natural switching, the experimental switching studies have shown that compensation though significant is on average only partial (about 50 per cent) and that blood nicotine and COHb levels are lower after switching, albeit proportionately less than the reduction in yields [6-9]. The experimental switching studies have all been short-term, two weeks or less. It is not clear whether this is the reason for the discrepancy. Would the subjects in the experimental switching studies, who compensated only partially in the short-term, have compensated fully if the studies had been prolonged? This important question is unfortunately not answered by two longer-term studies of mouth-level nicotine intake (based on butt analysis) [10, 111. Aside from their contradictory findings, mouth-level intake does not reflect the depth of inhalation and 'u a poor index of blood nicotine and COHb levels [ 12]. In this report we present the changes in blood nicotine, cotinine, COHb and thiocyanate (SCN ) levels of twelve smokers before and after switching to a low-tar, low-nicotine cigarette for ten weeks. We were particularly interested to explore the extent to which these smokers responded to this relatively long-term switch by inhaling more deeply to compensate for the lower yields or by adapting to a lower intake of smoke constituents. Subjects and Methods The subjects were recruited by advertising in local newspapers for.rolunteen to take part in a study of less harmful cigarettes. 66 subjects attended on the first occasion and 29 completed the three attendances required for collection of baseline data (see below ). These 29 were then randomly assigned to switch to one of two types of 'safer' cigarette for ten weeks. One group (N=15) were given a specially designed low-tar medium-nicotine cigarette. The other group (N=14) were given a conventional low-tar low-nicotine cigarette. This report is confined to those who were given the low-tar low-nicotine cigarette. Since we are concerned here with responses to switching to lower yield cigarettes, two of the subjects who were already smoking low-nicotine cigarettes (0.7 and 0.9 mg ) have been excluded leaving a final sample of 12 subjects (8 women and 4 men ). Details of the subjects and their usual brands of cigarette are shown in Table 1. The subjects attended on nine occasions over a period of 12 weeks. They attended in the afternoon or evening of a weekday, the time of attendance of individual subjects being similar (within 30 min:) on all occasions. On occasions I and 3 the subjects had been smoking their usual brand in their usual way. On occasion 2 they had been smoking a popular commercial brand of middle-tar, medium-nicotine King Size (84 mm) filter cigarette (Tar 19.0 mg, Nicotine 1.3 mg and CO 20 mg) throughout the day. On subsequent occasions they had been smoking the experimental low-tar, low-nicotine cigarette which was also King Size (84 mm) and filter tipped (Tar 10.9 mg, Niootine 0.7 mg, CO 12.9 mg). There were I

I I I I I I I I I I I I I I I I I I Iang-Tirm Suritdting to Low-Tar Loro-Niuotint Cigarcttcs Table I Details oC the 12 subjects and their usual smoking habits 147 Usual 'Usual Brand Standard yields (mg) Sex & Age of subjects Consumption (cigs/day) Site Length (mm ) Tar Nicotine CO 1. F 35 35 SasaIl 65 19 1.3 15.7 2. F 56 50 Regular 69 19 1.5 19.0 3. M 39 30 Small 66 16 1.1 13.3 4. F 50 40 King Size 84 18 1.3 18.8 5. Xi 41 35 Small 66 19 1.2 18.3 6. F 31 30 Small 66 19 1.2 18.3 7. tii 46 30 King Size 84 18 1.5 18.3 8. F 43 60 Long Size 79 12 1.3 11.5 9. hi 27 40 King Size 84 18 1.5 18.3 10. F 26 20 King Size 83 18 1.4 18.0 11. F 27 40 King Size 84 18 1.5 18.3 12. F 26 45 King Size 84 15 1.1 15.6 Mean 37.3 37.9 76.2 17.4 1.33 17.0 SD 10.2 10.5 8.8 2.1 15 2.4 ' AJD cigarsttes were 6ltertipped. about 3 weeks, on average, between occasions I and 2, and about 7 weeks between occasions 2 and 3. After switching to the experimental low-tar low-nicotine cigarette, subjects attended after 2 days and then at 2-week intervals for 10 weeks. The experimental cigarettes were supplied by British American Tobacco Company (BAT). They were in plain white packs of twenty, labelled with a code number only, with no indication that they were low-tar low-nicotine cigarettes. To provide some, but not too much, encouragement to participate in the study the experimental cigarettes were supplied at roughly 30 per cent discount on the cost of the subjects' usual brands. Travel expenses were also paid but there were no other financial inducements. Prior to each attendance subjecu kept a record of their cigarette consumption that day up to the time of arrival as well as.a sample of their cigarette butu. On arrival they smoked a cigarette on a puff analyser (13). On some, but not all, occasions a venous blood sample was taken two minutes after the cigarette was completed. The'blood samples were taken on the first three base-line occasions and then at 2 days, 8 weeks and 10 weeks after switching to the experimental low-tar, low-nicotine cigarette. They were analysed br COHb using an IL 182 CO-oximeter [ 141 and plasma nicotine was measured by gas chromatograph j 15). The nicotine content of the cigarette butts was measured by BAT. Subjective ratings of the acceptability of the cigarettes were taken on all occasions. These included ratings of satisfaction, taste, ease of puffing, missing usual brand, diSficulty keeping on the experimental cigarette and perceived ability to change permanently to it. The main statistical analyses were orthogonal single degree of fseedom comparisons (16), firstly between the usual brand and the control middle-tar brand and secondly between the usual brand and the low-tar low-nicotine cigarette. Other comparisons were made using t-tests. 0 i

I I I I I I I I I I I I I I I I I I 148 Russcll, Sytto>'t, Iyer, FeyerebeAd axd Vesey Resulta The various measures of smoke exposure (cigarette consumption, mouth-level intake and blood measures ) of the 12 subjects are shown in Table 2 and Figure 1. It can be seen by the points in the figure and the SDs in Table 2 that there was wide variation between individual subjects. The average levels, however, were very similar on the two occasions they were tested when smoking their usual brands. Likewise, after switching to the control middle-tar brand they were a2w similar, and apart from the tendency to puff less nicotine from the control middle-tar brand than from their usual brands (mouth-level intake 0.9 mg vs 1.14 mg, F-4.97, df 1/11, p<.05 ), there were no significant changes in the various measures of smoke intake after switching to the middle-tar brand. On switching to the low-tar low-nicotine cigarette, however, there were substantial reductions in some of the measures of exposure as described below. Table 2 Cigarette consumption, mouth-level intake of nicotine and blood concentrations of vari- ous smoke components of 12 stnokers when smoking their usual brands and after switching to a middle tar brand br one day and a low-tar low-nicotine cigarette for up to ten weeks. (Means and SDs ). Middk T Low-tar, Low-nicotine brand Usual Variable Brand ar Brand Usual Brand 2 days 8 weeks 10 weeks Cigarette consumption 23.8 20.2 20.2 21.2 23.2 23.9 ( 8.1) ( 7.2) ( 7.0) ( 7.0) (13.4) ( 8.9) Mouth-kvel nicotine intake (mg/cig) Smoked in laboratory 1.61 - - 0.83' - 0.74' ( .48) ( .24) ( .17) Smoked naturally 1.14 0.90' - 0.64` - 0.60= ( .36) ( .23) ( .22) ( .15) Plasma nicotine (ng/ml) 32.4 35.2 33.2 20.8' 22.T 22.8' (12.3) (15.4) (14.3) ( 8.4) ( 7.7) ( 6.5) Plasma eotinine (ns/ml ) 350 - - - - 246' (140) (134) COHb (!i ) 7.4 8.0 7.0 6.9 7.6 7.8 ( 2.0) ( 2.3) ( 2.4) ( 1.9) ( 1.7) ( 1.6) Piaasta thioeyaaate (t~enol/1) 150 156 140 142 155 167 (36) (35) (S1) (32) (36) (75) Note: The eeder of the oolttttunt refleets the atkr of switc4iug te vatiour braads (iee tsn ). Ci4uette eoesump- doe refers eot to a fnll day's ooawmptioa but a the stanber of ciptetta smoted up w the ti[ne of attendance. >tlout6-kvel istake is aa escimate of tbe dou of aieotiee pWSed iato the nouth. It was calculated fran the butt aicatine metent and the GIser edieieaey. Data aa plawaa catieine and .oab-ievd intalte were incomplete (.re oe:t ).o tbttt tye valnes for sval beaed are the eeana ot tLe two ." brand oecasioos aad the values ibown at 10 weeks oe the low tar Ixaed aee the taeam for i and 10.reeb. The steses tor'tuma cacmiae tepeseet onl) 8 ef t6e 12 snbjeet: as do thoa for tie nouthlevet iuakr daiootine 2 days alter switeA* to the low•4ar. bw-aieatieie beand. The kvelt otwtistiea! siBaiCicasee ttkr a mmpariteaawith tLr dsual brand; `<A5,' p<.0I .' p<.001. I

i Long-Term Swikleing to Low-Tar Lom-Nicotine Cigarettes 149 ~ ~ 60 50 40 30 20 1 t 1 1 ~ ~ 10 8 6 4 Plasma Nicotine (ng/ml) 10 U1 --1 Brand COHb (%) Low Tar Brand Plasma Cotinine (ng/ml) 600 500 400 300 200 1001 1 Usual Brand 12 r 2.0 r . _ ~ LJ_ Z Usual Brand - Low Tsr Brs+x! .. Mouth-level Nicotine Intake (mg/cig) . 1.6 1.2 0.8 _ _ .. 0.4 0 i Low Tar Brand i Uswl Low TK Brand &" Fipre 1 Messarm af anoke intake wben tmokiag trwl braad: and after nvitc3tisK to a low-car low-aicotiae cigarette br 8-10 weehs. EAcb paiat is the meaa oi cwo oaasio~n. ~

I I I I I I I I I I I I I I I I 150 Russell, Sutton, lya, Feyerabrnd and Vesey Cigarette con,rioreption. There was no significant change in cigarette consummptic>n on •witching frotnthe usual brands tfl the low-tar cigarette !F=0.45, df 1/ 11, NS, Table 2). Mouth-level nicotine iatahe. This is an estimate of the amount ofnicotine puffed into the smoker's mouth and is calculated from the butt nicotine content and the efficiency of the filter. When smoking their usual brands under natural conditions outside the laboratory the subjects puffed an average of only 1.14 mg nicotine per cigarette (Table 2). This is 14.3 % less than the 1.33 mg average of the machine-smoked nicotine yields, but this difference is not significant (t=1.8, df 11). the month-kwel 4nske of nicotine was significantly ~reduced after switching to the l~ow-tar low-nicotine cigarette when smoking in the laboratory and under natural conditions (Table 2, Fig. 1). The amount of nicotine the smokers pu$'ed from the low-tar low-nicotine cigarette was similar to its standard machine-smoked yield of 0.7 mg nicotine. The amounts ofnicotine puffed from the cigarettes were consistently higher when smoked in the laboratory compared to natural conditions, the increases averaging 31.3 % for the three comparisons shown in Table 2. There was a systematic decline in this di8'erence as subjects became more fimiliar with the laboratory procedure, the laboratory smoked level being 41 % higher for the usual brand (p<.001), 30 % higher for the low-tar cigarette at two days (p<.025 ) and 23 % higher for the low-tar cigarette at 8-10 weeks (p<.05 ). Plasma sicotnu. The average plasma nicotine concentrations on the two occasions of smoking the usual brands were very similar (t=0.2, Table 2). They were also fairly consistent within subjects (r=.88). The average level was also similar when smoking the control middle-tar cigarette which had a similar nicotine yield to the average of the usual brands (1.3 vs 1.33 mg). On switching to the bw-tar low-atcotine cigirette, plasma nicotine kvels' were reduced subniatiilly (F=22.7, df 1/11, p<.001). There was no significant difference between the short-term reduction of 37 % at 2 days and the longer term reduction of 30 % at 8 and 10 weeks (F=2.3, df 1/11, NS). These reductions were, however, less than the 47% reduction in machine-smoked yields from 1.33 mg to 0.7 mg nicotine (Figure 2). Plesnia cotiaiu. As a main metabolite of nicotine, plasma cotinine provides a useful check on the plasma nicotine data. Another advantage is its longer half-life in plasma. Whereas plasma nicotine levels arr determined predominantly by nicotine intake from the cigarette smoked just before blood sampling, plasma cotinine reflects the nicotine intake from all the cigarettes smoked during the day and to some extent the preceding day. It is only recently that we have developed this measure so that it was done only in tbose samples that had sufficient plasma left over for storage. The aimited data fi+em eight ofthe subjects do, however, support the pluasa niootine data in showing an idenatic:l and aatisticalty o6ni5unt redtsctioo d301i at 8 and 10 weeeits a&er rvrkehing so the bw-mar low4iootme ciguetoe (t=6.3, df 7, p<.001). COHb. There were no significant changes or trends in COHb on switching to the control middle-tar cigarette or to the low-tar low-nicotine cigarette despite the fact that the machine-smoked CO yield of the latter was 24 % bwer than the mean CO I

I I I I I I I I I I I I I I Loag-Terrn Switeking to Low-Tar Low-Nicotine Cigarettes 151 yield of the subjects' usual brands. There was some consistency within subjects on the two occasions they were smoking their usual brands (r=.73). Plasma thiocyanate. Tlie very long half-life in plasma (2 weeks) makes this measure inappropriate,for assessing short-term changes in smoking. There was a fair degree ofconsistency within subjects for the two occasions on their usual cigarettes (r=.76), but there were no significant changes after switching to low-tar low-nicotine cigarettes. Direct comparison of the means of the two occasions on the usual brands with the means of the 8 and 10 week values on the low-tar low-nicotine cigarette showed that they were not significantly different (145 vs 161 umol/l, F=1.1, dfl / 10, NS). Talbk 3 Measures of pu6irtg behaviour when smoking usual brands and after switching to a middle tar brand for 1 day and a low-ur low-nicotine brand for up to 10 weeks. (Means ) Middle Low-tar, bw-nieotitte brand Variable Usual Brand tar Brand 2 days 8 weeks 10 weeks Butt kttgth (mm) 27.8 30.7' 28.0 28.5 27.7 Number of pufTs 14.7 15.3' 14.6 16.8' 17.3' Inter-puff interval (sec) 25.6 28.4 28.3 25.4 26.2 Total time to smoke cig. (min) 5.4 6.2 6.3 6.2 6.4 Puff duration (sec) 2.3 2.3 2.5' 2.3 2.3 PuBpre:sure (crnWG per sec) 28.0 34.2` 30.7' 27.8 28.0 Puff volume (ml) 39.8 38.8 38.5 36.7 35.8 Total volume puffed (tn1) 547 569 5,49 573 579 Note: Due to missuta data tlr values for the tusual bnnd an the meatu of the two oeeations oa whieb uswl brands were tested, and the values for the asiddk-tar brand are the means of only e of the 12 subjects. The levels of statistical tisnificance rekr to compari.m with the usual brsad; ' p<.05,' p<.Ol. Pu„d-sag Behaniour. Although the subjects were dearly puffing their usual brands faster than the standard machine-smoked rate of one 35 ml puff of 2 sec. duration per min, there were no significsnt changes in puff volume, oeesl volume puffed, puff rate and time taken to :moice a cigarette after switching to the eontrol middle-tar brand and the low-tar lowvdootiae cigarette (Table 3). Subjects left longer butts when smoking the control middle-tar brand and tbere were sig- nificant increases in puff pressure and the number of puffa. After switching to the low-tar low-nicotine cigarette, pu>g' volume was maintained in the short term (2 days) by increasing puff pressure and duration, whereas in the longer term the number of pu$s was increased. These changes in puffing behaviour could have been due to differences in draw resistances of tbe cigarettes but we had no data on the draw resistances of the usual brands. Wbat is 'siaportaat, bowever, is the Lck d'signifiaat ebana U puff voicme aod soul volume puSed after switching to the low-tar bw-a>~oocioe titsrewe- Aasptability. Subjects rated the control middle-tar beand as less good to taste and less easy to puff than their usual brand (Table 4). The bw-tar law-oiootine I

I I I I I I I I 152 Ruuell, Sutton, I,yer, Fryercbend and Vesey TaMe 4 Subjective ratings of acceptability of usual brand, a control middle tar brand and a bw-tar iow-aicotine brand. (Means of 12 subjects) Middle Low-tsr low-nicotine brand Usual tar Usual Variable Brand Brand Brand 2 days 8 weeks 10 weeks Satisfaction (100=very) 86 62 79 71 81 82 Taste (100-very good ) 82 68' 83 72 76 73 Too strong (100-much) 25 32 26 25 13 18 Too weak (100-much) 13 24 22 34' 2T 26' Ease of puffing (100=hard) 17 30' 14 33' 19 22 DitLculry keeping on test cig. (100=very ) - 10 - 10 5 5 Missed usual brand (100-a lot) - 17 - 26 11 11 • Cou1d stay on test cigarette permanently - 1.75 - 1.67 1.0 1.0 for 8 weeks - 1.5 - 1.5 1.0 1.0 Number of subjects who prefetxd test cigarette - 1 - 2 10 l0 •Definitely - 1, probably - 2, pnobably aot - 3, definitely not - 4. The kvelt of statistial sjnifinnee refer to coatpariaons with the usual brand;' pt.05, i p<.p1. eigarette was rated as a little too weak but it was compared very favourabty with ~ the control middle-tar brand in the short tam, and in the long term is rated as well as the usual brand apart from being too weak. Indeed, in the long term it was sctually prekrred to the usual brand and all twelve subjects thought that they . eould cbange to it permanently. ,i Degree of compeasad'oa. Oa awitehing to the bw-tu lowrnicadne ei6arette the seducti+oa in nauth-ievel nicotine intake, plasma nicotine and cotinine were (as ~ mentioned above and shown in Fig. I and Table 2) substantial, consisteat and .tatistically si6ni6cant The reductions in plasma nicotine and cotinine were, however, proportionally less than the reduction in standard machine-smoked nic- ' otine deliveries indicating that the subjects made some compensatory increase in /!!! inhalation. Some idea of the degree of compensation is gained by the amount that the observed values were raised above the levels that would be expected if there was no compensatory change. This is shown in Figure 2:~it caa be aeea that there a wu uo.ownpenatioa fn terms ef mouth-leslel intake, but piauna aiootine and , .ootinisne values showed partial oompenation in the amo®t of ntoatine inhoed. In contrast, the COHb data showed complete compensation, the levels on the ,~ low-tar low-nicotine cigarette being similar to those on the usual brands. I Expressing compensation in perantage terms (Table 5) makes it possible to compare the degree of compensation shown by diSerent measures of smoke intake. Other calculations are shown for the compensatory increase in smoke ' intake and the theoretical increase that would be necessary, given the reduction in machine-smoked deliveries, to achieve complete campensation. This latter measure is less i'or COHb (31.6 %) than i'or nicotine and its metabolite cotinine (88.7 %) because the reduction in machine-anoked CO yield of the low-tar low- ~ ~ nicotine cigarette was less than the reduction in nicotine yield. The compensatory ~ ~a. © ~ 00

I I I I I 1 I I I I I I I I I I Loag-Tenn Switching to Low-Tar Low-Nicotine Cigarctka 153 Tsjbk 5 Calculations br compensatory changes in smoking behaviour after changing to bw-tu, low-nicotine cigarettes for 8-10 weeks. (Means of 12 subjects). The machine-smoked ratio br ootinine is based on the nicotine yield. Mouth-kvd nicotine intake Plasma nicotine Plasma eotinine COHb AlacAruvrntaErd yu/dr Usual cigarette 1.33 mg 1.33 mg 1.33 mg 17.0 mg Low-tar cigarette 0.7 mg 0.7 mg 0.7 mg 12.9 mg LT/ustul (a) .53 .53 .53 .76 Memmed istlmEe On usual cigarette 1.14 mg 32.8 nglml 350 ng/ml 7.2 % On Low tar cigarette 0.60 mg 22.8 ng/ml 246 ng/ml 7.7 % LT/usual (b) .53 .70 .70 1.07 Co+espt+tutory isk-reau a Moke raemtr (%) b -IJx 100 (c) 0.0% a Issrreau i+r nwo,Er ottate aaruory for cmplete eenrpeacorron 1- a x 100 (d) 88•7% ~ ] 2.1% 8.7% 2.1% 8.7% 0.8% 1.6% a lktrer of c.wtpruattan (%) d x 100 0.0 % 36.1 % 36.1 % 130 % Comparisons of the usual and IoM tar, low-nicotirx brands are based on the mearu of the two oeeasions on the usual brand and the means of the attendances aher 8weeks and 10 weeks on the fowaar, bw-nicotiee eigarette. The cotinine values are the means of only 8 of the 12 subjeets (see te:t). increase in smoke intake was nil at mouth-level, but the blood measures show the increase in inhalation to have been somewhere between 32.1 % and 40.8 %. The cyanide yields of the usual brands of cigarette were not available so that the plasma thiocyanate levels could not be used to estimate the corresponding measures of compensation. Estimatcs of rcduction as tar attake. Tar intake cannot be measured directly. How- ever, it is possible to estimate di>8'ertaces in tar intake Erom different brands of cigarette by using blood levels of an appropriate marker and relating these to the ratio of the machine-smoked yields of tar and the marker to give a tar intake index for comparison between brands. Such an estimate assumes that the ratios for human smoking are directly proportional to the machine-smoked ratios and that tar deposition is directly proportional to absorption of the marker. Using nicotine as the marker, for example, a tar intake index for a particular cigarette can be obtained as follows: Tar intake index = plasma nicotine x Tar/Nicotine ratio With this formula the mean tar intake index for the twelve subjects on their usual brand was 32.8 x 13.1 - 430 and on the bw-tar low-nicotine brand (after 8-1Q weeks) it was 22.8 X 15.6 - 356, i.e. a reduction of 17.2%. Using similar calculations the reduction based on piasma cotinine was 16.3 % and that based I

I I I I I I I I I 154 Russell, Sutton, Iw, Ft,yerabend and Vcsry 40 COMpENSATION AT 2 DAYS APD 8-10 WEEKS Plasma Nkotine (npJml) PlesmsCotnine (npJml) , 400, ao- 20 10, 0 . . 7 a0o-~ 200 100 0 tJkr! LT lT LWAI lT 2days i-10 •-10 M.ks .rphs COHb (y.) lAOtlth-IeMN NlCditE Irrtihe, rt6tur61 ttnoKrp (mpkp) :-; 1.Z ^ 4. ~-i r-• !~ 6...-1 081 0-41~ ' I I I I I I I 0 ` . 0 LUsuM LT lT Urrl lT lT 2days i-10 ld"s i-10 .w"s wefts F#;ue 2 Compensation at 2 days and 8-10 weeks after changing from usual brands to a low-tar bwqiooti»e (LT) cigarette (meaas of 12 su*cts, 8 in the ease of cotinine). The shaded areas tboa• the amount of contpen:atioa and repreieat the diBereate between the obarved leveh (npper point ) and 'expected' levels (lower poiat). Tbe 'excepted' leveh are tbou predicted 5om the tedwtiest in machine-anoked yield assuming that tltere was so e5atp in tutyokiai patteta. The broken lirrcs tspee.eat the kveit of tbe ttaadaze! m.etiioeytnoked yidds. SqaiBcsna of ob.erved-espected dif- fiereaca; ••p<Al, •••Pt.001. on COHb (after :ubtractiag 1.0% br background level in non-smokers) it was 11%. If these estimates an eorrect in showing a reduction in tar intake after switch- ing to the low-tar low-nicotine cigarettes this would indicate that, as with nic- otine, titruioa of tar intake was also incompkte. Using calculations shown in Tabk 5, the estimated degree of compensation of tar intake was 54% using nic- atine and cotinine as markers and 68% using CO as the marker. , I