Tobacco Documents Online

1` c tae, Wv G- b tv b~ RfGIUTORY IOXK"OU)G\'ASO.rHARMH(1D1iO613:. 1I10-1_o.(19a3( Smoker Intake from Cigarettes in the 1i-nng Federal Trade Commission Tar Class Gto, B. GoRI AND CoturF.utus 3i LnNct;t The FiankJln lnttirm+4 1330 fenwiEh taae. Siher Sprinl AYaj,>iand 209 10 Rereired lderuory• 13. 19d3' GLarexte yields measured by the atandand analyticg poceduta of'tde U. S. Federal Trade, Commissioo (FTC) may not be sutfiatady, infotma(ive to smokas.orlow-yidd brands tieause . actaal intake is likely to depend ttuinly on the agpestivenea of petsoaat behavior. This uudyy determined intake in smokess of I-a+t FTC tar dass ciiatettta. as tfxy swiuded IbtnndS. t'listaa counine terets used as a1 marker of intake, spanned I over a sita'1at tanfe of values koRa non- deteatable to about 800 niliol in all brands teaed! Pharntaookiaetic consideratioas sugtcst that smokers of thae brands--zs aVoup-iatake nieotine in excess of posted FTC values: However. mean values aams stt+okasfar eacb txand aswdt asitbe brand diferenoes io indiiriduakstaokcn, •Kte eiosely ptopotticnal to tAe analytica!',diRttenoes of'FTC nicotine yidds far each brand smoked Tbus standard anafytinl!valiks may not predict absolute intake of smoke.' but'ttwy appear to infonm, abaut 11w relative intake smokers can expert from difren:nt txands in the t- mi FTC tar dzst. IMLRODUCTION Numerous studies of cigarette smoking indicate that the intensity of effects depends on the amounts inhaltd' (Surgeon General, 198 1). On this : basis, cigarettes of ' low smoke . yield have been 1 advocated as being less hazardous (Gori and I Bock: 1'980): In the United States, cigarette yields are officially ranked using an analytical pro- cedure adopted by the Federal Trade Commission (FTC) with wide concurrence (Pillsbury er'aC. 1969j. Similar methods are used in other countries_ Lately„concernss over the differences of individuali smoking behavior have questioned whether such standard methodology is relevant in providing the consumer with a realistic measure of expected intake ()f.oYlo*vski et al:. 1982; Russell el al:. 1982). The particular im- plication has been that variance may be more pronounce& for smokers of low-yield cigarettes. The present stvdy set forth to measure the range of intake in individual'smokers of three popular U: S_ cigarette brands, advertised as nominally yielding t-mg FTC' tar. It' also addtessed the question iof how smokers ncspottded to small differences in FTC smoke yidds as thcy switrhed among these brand+s At the present state of the art; smoke intake in man can be measured directly through suitable physiologic markers. The use of carbon monoxide (CO) for 6ddl 110

SMOKERS tKTA'KE FROM CIGARETTIES III studies is questionable. because smokers are exposed to CO from environmental sources and also from cndogcnous physiologic processes (Ashton et al.. 1!9'81, Sjos- ttand! 11951); Nicotine is a specific marker of tobacxo smoke, although it u present in insignificant amounu in tomatoes, peppers, and' eggplant. Its terminal! half life is over 100 min in, the bloodstream (iAtmitagc et al.. 1975; Bcnowitz ei al.. 1'9'82; Cohen and Roe, 1981; Gritz,tt al.. 198'J'; Rosenbc.rg a al:. 1980), and likely to make its determination sensitive to short-term samplings variables and' dependent on the last two or three cigarettes smokedi This is of concern because individuals smoke with d.ie-ercnt in- tensity at diH'cnent timesi ldeally, a marker, should give an average indication of'the cumulative results of discrete intake episodes over a suitable period of time, and not simply a momentary experience. Cotinine is a principxl and direct metabolite of nicotine, from which it arises at predictable and rates (Coben and I Roe. 1'981'; ,Gritz er al.. 1981; Langone et al.. 1973: Matsukun et al.. 1979; Rosenberg a al.. 1980). This correspondence is confirmed by a high linear correlation of nicotine and oootininc plasma levels, as reported later in this study and by others (Gritz ef al.. 198'1'). A terminal' hall=lik of about 20 hr. (Cohen and Roe;,198'1; Langone et aL. 1973; Matsukuta er al.. 1979) only requires that subjects be sampJed'Iduring~latt afternoon1 and at fixed days duting the week, to synchronize , weekly and circadian steady-state patterns (Gritz et al.. 1981). On this account,,cotiaine is a better indicator of overall cumulative smoke intake. SigniBcantly, nicotine is delivered in close physical i association with tar partides (George and Keith, 1'967): Only a negligible amount is delivered in the vapor pltase, and' at the averagepH of cigarette smoke most of it is ttot absorbed in the upper respiratory tract but rather in the lung, where it is rapidly transferred to the blood- stream (Armitage, 1973)j Becattse of tlus, nicotine and'ootinine are alio valid indi- cators of tar, intake, once the ratio of tnr to nicotine of a cigarette's smoke is known. The experiment was designed'so: that it oouldibe analyzed as a sequential one, in a selGmatching design where individual smokers provide their own control: MATERIALS AND METHODS Cigareues: The study utilized three commercial brands in the 1-mg FTC tar class, purchased from commercial distributors: Each brand came from the same production batch„to minimize input variance. Specific analytical data are summarized in Table l. Analysis of cigaretic yields was pcrfotmed!according to FTC proccdures (Pillsbury et al. l'969), utilizing 400 cigarettes per brand. Subjects. Subjects were selected who customarily smoked either brand A or brand B yielding ll tng, nominal FTC tar. The subjects (1117 men and 171 women) were approached randomly in shopping malls and through notices in community >D.eMS- papers in each of five cities. Qualified respondcnts were at Ica.st'. 21 years of age, had been smoking at ltast 20 ,cig.arettes daily for at least 3' months, and engagcd i in no other smoking-related practices (cigars, pipes; snuB; chewing tobacco. nontobacco smoking products). Only individuals in good health, nonprcgnant. and'6 under no medication were necruitc,d. Subjncts were dropped from the study ifdisaasaor medical treatment intervcned: Individuals with alcoholic problems were also cxcludcd. Each subjectt was informcd at the outset that: participation in the study would rexquirc

11= GOR! A\D LYNCH TABLE I ANAL1l7CJA'L DATM oF GGAR FTR~ES' TCSTED' Wiean' and SD` Brand Batch code . Tar . Nicaiac A iXL 0.9=0? 0.1'6i 0102' B 00Q'-5I 0:3' ± 0-2 0.10 : 0.02'. C ~EB' 0;6I! 0.2 0.J11'! 0.02' •/1D cipranes !5 mm kn6tA. fi1tuI soft tacL ' 1h linillipann perieiEarene: Four bundtud aprcnes per brand*iue tested_ smoking specific f-tng tar brands provided free of charge; they also were asked not to consciously change any of their customary smoking practices. Srhedu/e. On their first visit, the subjects:completed a brief qpestionnaire on their smoking,history and related factots, and provided a lOlmi sample of blood. Subjects were aliways sampled in the late afternoon on Wednesdays or Thursdays. Each respondent was given a week's supplly of hLqber own customary brand of cigarettes to smoke for the foBowing week, andi a cigarette tally sheet to record thee exact time for each cigarette smoked since the day preceding sampun& They were also prouvided with special omntainersto collect allicigarette.butts during the ta0ying pcriod i as objective evidence of oonsumption.ll';espondents reported on the following week and again provided blood samples, obtainedladditional cigarettes, an&turned in the tally sheets and oollected' butts At tftis point, smokers of brand A were given brand B and'vicc versa_ Respondents stayed on the alternate brand for 3' weeks, reporting to the test center weekly to provide blood samples, etc. Starting with the fifth visit'to one of the test otnters, all, respondents were given brand C cigarettes to smoke for 2 woeks, once again raeporting to the centa, each week for blood samples, etc. Ali' subjects received a nominal monetary compensation at each sampling; Nicotinr-catinine corrdtuuon, A separate group of 45 male and 41 female smokers of 1-mg tar cigarettes was recruited for vwalidating, plasma cotinine as a marker of plasma nicotine and ismok'e intake. These subjects had'lhc same general characteristics as the main group above„ but spieaal precautions were taken to ensure that daily steady-state levels of nicotine and aotinine were synchronized, as required by phar- macokinetic considerations and previous experiences (Gritz et al:. 1981). Subjects in this group habitually smoked atregular intervals d'uring ahedlay„and'espctially, during the 3 hr preceding the sampling, which occurred on a Thursday between 5.T00 and I 7:00 Ptrt: A control I group of 23 male and 26 female nonsmokers was also i recruited and sampled once for plasma nicotine and aotininc B1ood samples: B1bod samples were drawn by a certified technician;, plasma was obtained I by centrifugation and' ftoun without preservatives within 15 min of being dtawn: Samples werrblind eoded and storad~ frozen~at -20°C or below until assayedi. Cotinine and nicorine analysis. Cotinine and'nicotine were detcrrnined using pub- lished mcthods (Jacob er aL., 1980). Standard curvcs wcre constructed and repeated every 40 dcterminations; by using,thc internal standards N-ethylnornicotine for nic- otine, and h'{,2'-mcthoxycthyl) norcotininc fbr, cotininc_ The peak height correlation

SMOKERS IATAJCE FROM (7GJtRETTiES 1113, was linear over the range 25 to 60D ng/ml for cotininc, and 0-100 ng/ml for nicotine: Throughout the procodure, prerautionswere taken toavoid exogenouscontamination, (Feycrabend Iand Russell, 1980). As a quality control lcbeck, 178 plasma samples were selected randomly for reanalysis of cotininc values. The a verage of the original values was 246 ngJml, with a standard error of the mean (SEM) of 1'1.4 ng/ml. The duplicate values had an average of'258 ngfml, with SEM of 14.9 ngftnL The simple cosTClation cocfficYent' between the original and I duplicate values was 0.93. and the paired t test ans not statistiaalty significant (P > 0105)j For nicotine, 21 samples were seleaed at random for nanalysis. The average of the original values was 26.3 with SEM' 2_S' ng/ ml; the duplicate set bad an average of 2G.8 with SEM 2_3' ng/ml. The paired i test value was not statistically signi6cant, (P > 0.05). To ezdude possible interferences from other oompounda, a random group of sam- ples was analyzed by high-resolution gas chromatography and mass spectrometry (GS/MS). Ten plasma samples, previously analyzed as above, were processed ac- cording to the standard analytical methodology for the determination of cotinine. Prior to GC/MS analysis, the sample extracts were pooled and concentrated to a volume of 20 p~l. A 1y~1 aliquotwas then analyzad'by CClMS. The GCJMS conditionss were as follows. GC/lr(S: Finnigan 4000 AMU range: 35-450 Scan I speed: I scan/sec Mode: Electron Impact, 70 eV, positive ion Column: Fused s7ica„ 30 m X 0.312 mm Flow rate: 2 mlymin belium Liquid phase: DBS, 3A W' Scientific Temperature program: S0°C (4 min) to 280°C at 8°C/min Injector temperature: 260*C Injection volume: I µI, splitless No interfering mokcules were found. The calculated concentration of cotinine w-as. 594 ng/in) I by, GC/MS, which compared favorably with the 526 ng/ml measured by the GC analysis routinely used. RESULTS Subject cluvacteristics. General characteristics of the subjects entering the study are summarized in Table 2. The groups represented predominantly white collar oc- cupations an& housewives. Male-female dilfenenoes were rt:flected throughout the results, thus calling for separate analysis of the data. Pltrsmn cotinine. Tiable 3 summarizes weekly trends of mean values for plasma cotinine measurements separated by groups having homogeneous characteristics, i.e., by the brand' of cigarettes customarily smoked and by sex. Baseline values were averaged over the two initial measurements taken. Expected values were based on the FTC nicotine yield ratios for the brands smoked (Table f). No statistically sig- nificant di0'erznact were noted between observed and expected values in a self-match- ing paired t test analysis, except in the two instances given in Table 3. Although significant, these two deviations were small.

114 GORI AND Ll"NCH TABLE 2 RESPOYDEMT'C.7lAaACfERtSTIC3AT FIRST'EVTRy. Number of tespondeou A,mta4e No. of months smolun=, automary brand, Average No. , of,cFarcaa amokwday Averate a8e Cuaomuy txand! of aipratc M F M' F M F 1:t F Brand A'' 67 75 51 6.4 30.6 29:8' 353 ~ 38.4 (0.4) (0.4), (0J) (p.'9) (t.1) (12Y Brand B 50 96 10.8 10.6 301 27.7 39.1 39.1. (0.51 (0.4) (1.1) ' (0.7)', (a.7) (1.2) Nort. Standard iertor of the me.n in parentAeses. ' Braad A,bad been introduoad'oo the market about l01tnootlis prror to the Aady. Brand B bad heen on Ne market for tevcral years: Tbe results indicate that-after switching brands-habitiual smokers of btaod A reachod thc plastaa cotinine levels of habitual smokers of brand B and vice velsa. They also indicate a consistent stability of individual plasma cotinine levels week after woek. TABLE 3 Wmu.r Ttet3+mS oF RuswA Curm:vttrE nv SuuKts of DwnmFart Btwnos Subjeats Caestomanly aaolon8' Bi•and A Band B' Sea Initial wbjects Week M 67 F 78' AA 50! F' 96 Customary brred 1 334' 312' 228 201 Customaty lxand 2 308' 333 176 17® BaxGne overrl( 317' 328 206 189. (t319) (1'8A) (ll6•8) (10.9) Aheiaate Grand' 3 209' 207 305 291 Ahunate tirand' 4 235 210 342 286. Altetnate baa4" 5' 228 2ft 320 273 Alternate txand•'oworatll 222 207 322 284' (I13.9) (183) (24.3) (15.4) Brand C 6 240 212 202 192 Brand C 7 230 197' 238 179 Braod C.,oreQall 235 203 223 /85` (173) (14.0)1 (20.2) , (I1I1.6) Nora. Statiuid signifHnnor was determinad oo a self-matalun8', basis; paired r texts (twar tailedl, after adjusuns individlul'values for the number of daily aigarettcs eonsuawd: Valuns upressed Iaa means and SEM in nanograms per milliliter. "Brsnd A for brand B'smok'!rs and vioe versa- •'Sgni6ontly lower than expectod. P< 0.05: `Sipitirantly lower than expane4 P'< 0i0t.

SMOKERS IT.TAI:E FROM CIGA'RETTFS 115 In an overall assessment. Table 4 summarnzes the results for all respondents as they smoked the three brands tested. Beraust the level of cotinine partiallk• depends on the number of'cigarettts smoked'daily„a nonmalization was carried out for each, data ~point of each respondent, dividing individual plasma cotinine values by the number of dailyciga.rettes usead I prior to that sampling. Averagcs of these normalized values are given in Table 4'. The ratios of these aNerages agrueod we01 with the ratioss of tlie analytical nicotine yields of the cigarettes tested (Table 1'): The correspondence is especially dear, for the baseline mean plasma cotininc values for brands A and B, which are presumably, uundisturbed by experimental stress (Fig. I'). Maximum eie- ca,rded valua of plasma ootinine ane atso given in Table 4, indicating tltat eadt brand tested generated a similar range of values. The sample coefficicnt of skewness was calatlatad for each experimental set of vallus. In all casa the cocfficientwa's positiive, suggesting a slight dispersion of values to the right of the mean. No coefficient was significantly greater than zero, nor did any coeff'icicnt change significantly for any group. Number of dgarettes smoked daily. Table 5 summariees the average number of cigarettes smoked daily, separated by the brand customarily smoked and by sex. In all four groups of respondents, brand A cigarettes were smoked the Ieast: The men smoked the largcst numbers of cigarettes while on brand C'and the women smoked the largest numbers of dgalzttes while on brand I B. Tbe average change aas al,ways belbw, 110%. Paired t tests for each, individual experieaae revcaled no agnificant dit ferences at the 5% kvd; suggesting negligible compensation in the ntunba of aga- nettes smoked daity: INicori.u-cotfru,ne correlation and'nwssmoking csaurafs. Figure 2 gives data from the group of ' 86 subjects expressly sampled I to explore the correlation of plasma nicotine and aotinine. The linear correllation coefficient r= 0.84 was significant at P'< 0.001. In the emntrol group of nonsmokers, no subject had a detectable kvel'of plasma ootinine (25 ng/ml or grcater): Plasma nicotine in the males ranged from 0.5'5 to 4.8 ng/ml; with a mean of 1.9 and 0.3' ngftmi SEM. For the f,emales; the plasma nicotine ranged from nondetaatable to 7.7 ng/ml, with a mean of 1.5 and'03'ng/ml, SEM'. The male and female subjects combined had a mean plasma l nicotine level of' 1.7 and 0.3' ng/ml SFM. Observed'and exp¢rttd plasma catinine values: Available data on the pharmaco- kinetirs of nicotine and cotinine (Armitage er a!_, 1975; Benowitz, 1982; Benowitz. TABLE 4 SuMmARr oF PL4smA Connnra vA1.trES nv Smox>as oF DtFFaaFnrr BxAmm Btand A &aad B Bnnd C Mean piasma cotinine, ooml (SEM) 301 (10.1) 201 (82)' 208(8.2) Maximum value nxordo4,n&/,ml 833 659 E99 Mean agarettes smoked daily 29.0 30.5 31.1 Menn oormalixed'plaama cotinine, nt/mN' 10.11 6'.8 6:7' Rauo to brand C' 0.16 0.10 0!10' Ninaine myeiguntuc 0.18 0a ll', 0:10' ' For aab inespondent, plasma cotinioe,valua wcre divided by hisAer avua6c daily asaretu oonaumP tiou. Averages of overall results are reponad., 6 Ratioa of the normaliso4 plauna:ootinioo vallua udcinj Uie braod C value as 0:10.

116 CAR1 AND Ll'vCH Plasma. Cotpw+e nGlud a. W FTC'M"Kotiae mp P.r C'garNte. F1G. 1: Obser+ved'and upeciod bsseBae pisama aotiaioc valoes as a fuoatioa of'F7rC'niowine delivery of brands A and IIL er al:. 1982; Cohen and Roe, 198 1; Gritz et ct. 19811;3:angoneeral:. 1973; Matsukura er: al:. t!979; Rbsenberg, et vl.. 1980) ~ allow a nough estimate of the expected mean values of maximum plasma cotinine levels reached at virtualidaily steady-state con- ditions (C,,,.=), if the nicotine intake per cigarette were equivalent to the FTC(N), analytical values. The parameters utilized in the estimate were: eotinine half-life h = 20 hr(Bcnowitz, 1982; Langone et cL. 1973); volume of'distn'bution for ootinine Va, = 1000 m1/,kg (Benowiu, 1982); interval between cigarettes dt = 32: min (equiv- alent to 30 ~ci$arettes/day,k conversion rate of nicotine to cotinine k,, , = 0.7 (Cohen and Roe, 1981); and average weight of'subjects K' = 70 kg. The equation becomes (Curry, 1!980) _ FD(xN)' X L, ~.._ = R, x V4 h1 - (4.5)°`'")~- . Figure I indicates that -aa a goup-smokers of the low-yield brands tested tend I to intake nicotine in excess of posted FTC values. Plasma cotinine in smokers may not l TABLE 5 DA'ILY'CYGmRE7TE CONSUMlnoM IM'SMIOKERS ~OF t)1FFERENT BRANDS MCaflS and~.($EM). Customarry, brand Sex, Baselinc range N Brand A Brand B Brand C Brand A M 20-52 67 31(0.4) 3340.7) 34 (0_7). Brand 8 M, ' 2aSI 50 30140.8) 31(0.4) • 33 O1LtJ Brand A F 20-55 75 29(0.3) 31 (0_6) 30'(0.8). Brand 5 F 20-!7 % 27 (0:5) 28(0.2) 29 (0.6'1,

SMO):ERS tN7Ali.E FROMI C7GkRE77FS 117 Pl.stna CoAkBne aOa° ,wad f.fJaJ ..{f. w.f tm f WM . 1fo . am. fa9l no ' am Plssm Nioolww, np/ml FiG. 2 Plasma ootinine valucs at a fuoctiaa of piasma' aioo!tine vilues Maks and itmalat, 1leach true steady-state conditions on aoQount of the long balf-lilr thus, the actual difference between expected and'obscrvod values may, be wider t6an estimatbd: ~A1= though informative, ii'is aovessary to empbasizt the uncertain quantitative meaning of these conrllisions, owing to the diverse sources and natural varianoe of the pa- rameters utilized in this estimate, and'to the approximation i of assumptions. Plasma contnine as a fwuction of ci,garertes smoked 'dailp: Fgun: 3 ralates baseline plasma aotinine of femak smokers tb ava~abk da7y, nicotine (~N), defined by the FTC nicotine yield of t6e cigarette smoked (FTC(N)) and the number (CPD) of dgacenessmoked da7y. ADN'= f'TC (JV) X CPD: A posztive oorrelation.is dis+cermible; although the variance is large at r= 0.4i1. A similar distribution was noted for males, with r=-0.55. Both correlation ome6~cicnts' anere significantly greater than zero laoo aom t°lasm. CoNdn. 3Oa° . . . . .. . , , , .... , . .  . . . .. ~ . . . . ,.. ,. ~ . .~ .... ~ ... , , . ,. . :. . .. , ...~. . ~...~. , . ~ .. . f . s. ., . a . . ..as.s • •-a,o w.s. i. 1:30~. 2l0~. 1'b ~0~ ['SO &i0 ' Daily AraMb1e Niootine, nnp FIG:,3. Baseline pUuma cotininr u a,lunction of availabk daily niootine, de6nod by the FTC'nicatine mrlci`ararr and Ne dprcnes smored per diy: ADN - FTC (N) iX CPD. Maks

115 GARIi AT:D LYNCH (P'c0!01)1 The coefficients did not improve after adjusti!ng cotinine values for subjert weiRlat indicating thai behavioral' and sampling factors art the more iikeln sources of variance, metabolic factors probably being stable. as suggested b.• the high cor- rcJation between plasma nicotine and cotininc. IDISGUSShON' Plaarrtrt nicotirte aed mtininecturdation: The kinetics of nicotine intake, excretion, and metabolism are known to depend on differences in individual behavior, waght, lean body mass, urinary pH, ett., rrsulting, in mcasurably different plasma: nicotine and cotinine values in various individuals smoking the same cigarettes (Armitage et atl.. 197'5: Btnovitz et el.. 1982; Cohen and Roe. 19811; Gritz et al. 1981; Langone et al.. 1973; Matsukura et al:. 1'979; Rosenberg et al.., 1'980;' Russell et' al: 1982). Moreover, the metabolism of nicotine in the liver, its major site of'detoxification, is mediated by P450 microsomal enzymes that can be activated or depressed by a number'ofdietary components, alcohol, dbugs, pathologic, and physiologic conditions (Cohen and Roe 1981); However, two observations in this study suggest that these potential souroes of variance are relatively small in our sample, and at least within ~ the iange of plasm'a concentrations measured. One is the high correlation of plasma levels of nicotine and cotinine (Fig, 2)) which confirms previous findings (Gritz et al..,1'98'1). The other is the remarkable stability of individual plasma cotinine values week after week„while smoking the same cigarette (Table 3). The validity of these results is reinforced by the positive correlation of plasma cotinine levels and the number of cigarettes smoked daily (Fg.. 3)~, Intake. Any of the brands tested resulted in plasma; cotinine levels over a similar range of valUes. The virtually unchanging number of daily cigarettes smoked by individuals„andithe correlation of'plasma nicotine and cotinine valites„indicate thatt the same statement' is valid' for nicotine intake. ]n tutm, the physical association of nicotine and tar in cigarette smoke, and the nearly equal tar to nicotine ratio of the brands tested, suggest that simiiaroondusions are probably valid forur intake. How- ever, inferences on tar intake will have to be further verified by directl experimental evidence on the tar. to ~ niaotinc ratio in indi;vidual I smokers, which is bound to vary under different conditions of ventilation, puff volitme, pro6le, ditration, fre- quency, etc. Previous studies found that compensation under experimental switching conditions is usually only parrtial ('Russell l et al:. 1982). For the nominal' 1-mg FTC tar cigarettes tested in this study, small absolute differences in analytical cigarette yields did' not' change daily cigarette consumption, and resulted in plasma, cotinine levels propor- tional to such differences of FTC nicotine yields for the brands smoked (Fig. 1), suggesting no behavioral compensation uponi switching, For individual smokers, this proponionality, occurs within a limited segment of the overall range of observed values: high inhalcrs remaining high, and vice versa. This suggests that individual smokers of 1-mg tar cigarettes may find satiation at different lcveRof nicotine and smoke i'ntake. probably reflecting, the influence of exertion and effort in extracting smoke from low-yield brands. For equal exertion, smokers may attain nicotinc intakes roughly proportional to the cigarette yields_ Regardless of the proportionality of plasma ~

SMOKERS IA.TA'Y.E FROM CIGARETTES 119 aotinine and FTC'nicotine forditfetrnt brands, it is also apparent'.that most!smokers of1oM•-.uld cigat ettes compettsatt upward, and extract nicotine-and'hence smoke- in excess of what the FTC yields of individual btands imply (Fig. 1). CONCLUSIONS ' This study suggests that FTC'values for low-yield agarette brands understate the actual intake of avcrage smokets However. they appear to offer a valid representation of the relative intake thaindividual smokers can expect from the 1-mg FTC tar class agarettes tested. and probably from similar brands in that class. Since FTC' valites should be interpreted as liaving relative rather than absolute signi6cantae. they remain valid in ranking the expected' mean: relative intake from cigarettes in the 1=mg FTC tar class. Tbe meaning of FTC values in ranking higher- yield'cigarettes is being investigated in separate studies. ACKNOWLEDGMENTS The authorz ate pateful far the assistance of Lawen 8. Levetoo;,MS„ FtrokJia lnstitute, St7.a Sptii& MarylaaQ who coordiouad!the ovaraC lopstic>; Deborah A: ScAoeida: MS. Eltitk 1wid8einc, Atlaota, Geagia. who wperrrisod the tewitut8 of roliuttars and sample oolleaioq RooerA. Novak PtlD, Bor, tiston Resnafi labootories,l trc. Temple Hills. Marylaad,' wbo pafortned the analyd3 of pimna oolininC Neal L Beno.+itz. MD, and Pcyton Jacob. III, Ph.D:, (3inical Pharmacoio8y Unit. San Francirco Cxnanl Hospital. San Francisoo;.dw pufotmed the analysa af'ttlasma aiootinr~,and'The Brown and Wil(iamson Tobacco Corporation lnc., Louts>nik. Kxntucky, wha~e the tar and niaaine yield determiaations we:e pvfannod for the (ixancks used'in this study. The autltors also ttunk 1. Van Reswm. Prafeaar of 1'Aat- maodop•, Uni.+asity of Nijme+8en School of Medioine, The PktlKrlanQc T. D. Darby. Profmor of Phar- maooloey. Univrrsity of5outh C3rolina:,Srliool ofMedicine, Columbia. South Carolina: and'Proftssor L Denli, Unimsity of Basd Medical School. Switier(and, for advia on the pharmacokinctic nepuirements of study desi8n and lopstia. REFERENCES Astinor{ H- STOiPNEY, R, Atm TAOt.tr9DN, W, 1. (/981); Should intake ofcarbon monoxide be used as a 8uide to intake ofatbersmokroonstituents? Brit Ale+d J: 2182. 10-13. AwmrrAGE,,A. K(19731. Soene reoent observatioaa relating to the,absorption of'oieaine from tobacco smoke- ln Smoki)V Beharior.- AUoti.rs and Incnuivet (W. L Dunn; edJ, Wiastun, Washin8ton, D: C A'twtrAGE, A. K., Dota;>ner, C. T., GEORGE. C. F, HousE.uAN. T. H., t.EwAs, P: J. Arm TuaNER. ID M. (1975): ADsofption and metabolism of oicotine from ei8aretta. Brir: AiltQ: J_ 4, 31'3-31I6. HerqvmrZ N. L U1!982).' Ptrsono! C6.nrnunicaiiort. lb ptess. t3>:rrowrrz; N. L„JAcDn, P., Jor+FS: R. T„Ar+m Rostavs6RG, J. (1982): laterindi ridual variatiility in the metabolism i and cardiovascular dfara ofinicotine in man. J. P1Sartnaarol f.ap: Ther. 221, 368-372' Cotttal A. 1.. , A'ND RoE, F. J. (198 t). Monograph on the Pharrnaeob8y and ! Toxicolosy of Niaotine., Ocnaaionai'Papr.I. Tobacco Adnisory'Couno7. London., CLwer, S_ H. (1980). Dnra, Dirposiuion iard Pharr,waokinerict. B1iekwell,' Otfond. FEYEMnsrm, G. AND RtYSSEUt, Ml A: H: (1960): Assay of niootine in biological materiali: suurtts of contamination and their elimination. J. Pl+arrn. Plmsnacol' 32,' 178-18U. GEVrtGt', T. W., Awrp K,tSnH, C.. H: (N9G7)- Tlie selective 8ltration of tobacco smoke. In Tobacco and' Tobacco Smoldr (E L Wynder andlD. H!offman, ods:). Acndcmic Prass. Nc. Yorl'_ GoRi, G. S., ^Nm tlo[v.,F. G_, cds. (1980): A Sife Cig,anrnc?la Bonbuq• RrAon A'o 3' Cold Spnng Harbor hboratory, Cold Sannr Harbor; N. Y.