724 



140 



THE NEW ENCULHD JOURNAL OF MEOICINt 



ipcdfied in ih* FTC repon rf D««mber 1981.' Tbt p»po w« 

 removed and ttir lobaccs wtigbcd. NicouiK wu extncud by feaa- 

 ing (he tobacco is 30 ml of I ^ hydrochlonc add for oot bour ia a 

 boiling water bath. After the tuba >tood ovtnugfci, cbcy orre 

 miied, and IOO-/1I aliquou wen removed and diluted with 900 ;d cf 

 0.14 N aqueous ammonia. Niootioe coocentrationj were measared 

 by high-peHbnnaiice liquid chrocnaiography (Beekmao Modei 

 1 I2A pump with a Periii>-Elmer Modd LC-IS 6>cd mvckafik 

 detector >ei at 254 oro). A revene-phaae column was used (ABied 

 10 n C-8), with a mobile phaje clSi per cent methanol and U per 

 cent of an aqucoua lolulioa containing I per cent (vo(A«l) 2- 

 methoxyethylamine adjusted to a pH of 8 with acetic add, at a flow 

 rale of 3 ml per minute. The retention time ibr nicotine was I4J 

 minuta. .\n external aqueoui sundard of nicotiiK was ued, which 

 gave a linear standard curve over the entire range (C to 100 >ig per 

 milliUter) of observed coocentratioas. 



Two dgareties were measured from each pack. The nicotioe ooe- 

 centrations of dgarettes from the same pack differed on average by 

 i.O per cent 



To estimate nicotine intake, blood samples were uken from 272 

 subjects — 122 men and 150 women, with an average age of 37il 

 years (S D., range, 20 to 56) All sut^ecis were about to begin a 

 smoking-treatment program. Each subject was asked the brand and 

 number of dgarettes he or she had smoked in the previous 24 bourv 

 Blood samples (ot cotinine analysis were drawn at rwo diflereai 

 limes in two subgroups of patients. In Group I (n " 149), the sam- 

 ple was drawn late in the afiemoon, usually around 5 p m. (range. 

 100 to 6:30 pm.). In Group 2 (n = 123). the blood sample was 

 drawn at 8 a.m., after 8 to 12 hours of abstinence Cotmine was 

 measured by gas chromatography with nitrogen-phojphorus detet- 

 txm and a structural analogue of cofinine as the internal uandard.* 



Nicotine content was expressed as a percentage of tobacco weight 

 and also as the toul amount of nicotine in the airvouni of tobacco 

 usually burned in the FTC smoking-tnachine asaay. The relation 

 bnween the ccmcentratioo of nicotioe in tobacco, total nicotine in a 

 cigarette, and the nicotine deUvered by nuchiiK were measured by 

 linear regression. 



Dau on blood cotinine concentrations ia Croups I and 2 were 

 aiolyicd separately Univariate linear regression was used to meas- 

 ure the relation berween blood counine conccniraiion and age. 

 number of dgareites smoked per day. machine-determined nicouoe 

 >ield, and "daily doae* of nicoune (as estimated by the product of 

 number of dgarettei and the yield) Hierarchical mullipie liacar 

 regres]ioo'*wms used to examine the dScct of machine oicstax yieid 

 no biood cotinine coocentratio>, with coolroj far the tSeea of age, 

 sex, and number of cigarettes smoked per day. 



Results 



Cigarette tobatxo contained an average of 

 1 .57 ±0.33 (S.D.) per cent of nicotine. Neither the con- 

 centration of nicotine in the tobacco nor the total 

 amount of nicotine in the cigarette correlatetl positive- 

 ly with the FTC machine-delivery estimate (Fig. I). In 

 fact, among the cigarette brands sampled, there was a 

 significant inverse correlation between the nicotine 

 concentrations in the tobacco and the nicotine deliv- 

 ertd. 



Blood cotinine concentrations were significantly 

 correlated with the number of cigarettes smoked by 

 both groups (Fig. 2A). The correlation between ma- 

 chine-determined nicotine yield and blood counine 

 concentration was not significant in either group (Fig. 

 2B). The daily intake of nicotine (number of cigarettes 

 smoked per day rr.ultiplied by the nicotine yield of 

 each cigarette] correlated no better with cotinine con- 

 centration (r = 0.42 and r = 39 for Groups 1 and 2, 

 respectively) than with number of cigarettn alone. 

 Multiple regression analysis showed that combination 

 of the variables age, sex, and number of cigarettes per 



I 



Z 



< 



c 



o 



Figure 1 . Ucotine Contort of Otgar^lea. ■ Compared ** F 



OMeminMj Vatua* (Regnaaon Analyst). 

 "Viotaunti pkotr* per dgmot' danam total uncun a ■ 

 ■ne in the tangCi of cigareae-totoixo radanoked in Sw s 

 FTC smoUng-madwie assay. FHteen popiar co tnme mJ c 

 raOe tsrandi were 1 



day accounted ibr 21.2 per cent aad 23.3 per rrvi 

 the variance in blood cocinine ooncentraiica 

 Groupc I and 2, respectively. TW addition of 1 

 dune-delivered lucodiM yield added ooly 2.8 per < 

 and 5.0 per cent to the explained wiaoce in Groofi I 

 and 2, respectively. 



Tobacco from low-yield dgarencs does not 

 less nicotine than tobacco from higher-yield 

 This is probably well known to ^ tobacco 

 but to our knowledge, it has not bees disctissed 

 medical hterature. The main deterainants of wbrtks' 

 a cigarette has a low or a high yield ia machine irsiiaf 

 are the characteristics of its ventilatioo and bumio^ '^ 

 Both these variables are under the cootrol of the oj»- 

 Tette smoker and allow coosidcrabfe variation in mt»' 

 tine intake.' We also found that peo^ who wanted ■• 

 stt>p smoking and who mrere smobag low-yield ogs- 

 rettes did iK>t consume less nicodoe, as judgH bf 

 blocd cotinine concentration, than smokers of hjef>cr- 

 yieW cigarettes. 



Blood cotinine concentration is not a perfect iiurta 

 of nicotine consumption, because cf individual \^ 

 ability in both the conversion of mcoone to cotuui* 

 and the elimination rate of cotinine itself But we tu»« 

 no reason to think that there is an\ relation betwrrt 

 the rate or pattern of nicotine or cocinine metabolisa 



