Tobacco Documents Online

off the home pump, and completes the home diary. After work, on the early evening of Day 3, the subject returns to the test center with all of the material, complete,s the second visit questionnaire, provides a final saliva, and receives a $100 gratuity. The ETS samples are kept frozen, until they are analyzed by the Resurch and Development Department of RJ Reynolds (RJR), of Winston-Salem. RJR is responsible for all field operations and logistics, sample management, and all sample analyses. Bellomy Research is responsible for questionnaire design and field recruitment (through the aforementioned subcontractors), assistance with field operations, and coding of subject demographic data. Oak Ridge National Laboratory (ORNL) is responsible for overall study design and oversight, independent quality assurance of sampling and laboratory analyses, observation of and assistance with field operations, and data integralion, interpretation, and reporting. Briefly, two samples of each of three types are obtained from each subject: a home and work particle phase sample, a home and work vapor phase sample, and a saliva sample before and after the 24-hour sampling period. The saliva sample is analyzed for cotinine using radio-immunoassay techniques. The particle phase sample is analyzed for respirable suspended particulate;; [ 2.5 µm >] (RSP) by gravimetric measurement, and solanesol, fluorescing and ultraviolet absorl: ing particulate matter by high performance liquid chromatography. Vapor phase samples are analyzed for nicotine, myosmine, and 3-vinyl pyridine using gas chromatography. RESULTS As of the date of this report, field operations for 9 of 12 cities to be included in the study had been completed. Field operations for Buffalo (City No. 10) were scheduled to be completed November 5. Coded (magnetic tape) subject demographic data had been received for Cities 1- 6 from Bellomy Research. 'Copies of the non-coded work and home diaries have been received from Bellomy for Cities 1- 9. Sampling duration and pump flow rate data has been received for Cities 1- 9 from RJR. Smoke constituent sample mass data (ie. mass of constituent per sample) and salivary cotinine data has been received for City No. 1, Knoxville. All of the smoke constituent data has been corrected for blanks by ORNL, and only blank corrected data is provided in this report, ur less otherwise noted. The blank- corrected sample mass data is combined with the flow rate and sampling duration data by ORNL, in order to generate smoke constituent concentration data. In Knoxvillr, the pump timers on the personal sampling pumps were inadequately reliable, and sampling durations were taken from the self-reported information in the home and workplace diaries. Some general comments about subject recruiting are in order. In the first six cities, there has been a wide range in the number of calls required to recruit approximately 100 subjects in each city. For example, in Portland, ME, the success rate was nearly 100%. That is, nearly everyone contacted who passed the screening agreed to participate in the study. In contrast, nearly 5000 calls were required in San Antonio (City No. 3) to recruit the subjects. Because of the reduced presence of smoking in this country (both non-smoking individuals who live with smokers and workplaces where smoking is not severely restricted), it has been difficult in some cities to fill the population cells equally. For example, in City Nos. 5 & 6 (Boise and Seattle), only a total of 6 individuals (out of 191 subjects) lived in homes classified as smoking and worked in smoking locations. Based on concerns that the fraction of the study population in cells including a smoking venue may be too small on which to base definitive conclusions; it was decided to alter the recruiting practices in Cities 9 - 12 so as to fill each of the first three cells (those including a smoking environment) with approximately 25% of the overall subjects recruited for that city. It appears that this approach has been somewhat successful at increasing the mRAFr.• ,v*r# it e/# ask:wrnarwr&,d:ve,ixrro.wWO ad romrr;aV 89272850 2

study populations of Cells I - 3. However, perhaps as a result of some over-zealousness on the part of the local marketing research firms, there have been some problems as well. We have made some °~observations that suggest that the recruiting of some of the subjects was not conducted in a truly random fashion. In Cities 9 & 10 (Daytona Beach and Buffalo), it appeared that some of the subjects had been recruited through one or a few individuals who had provided names of friends, roommates, etc., who were in turn contacted by the local firm. In one case, one subject was under the impression that several of the subjects had been recruited through one individual who was su,;gesting that they all donate their $100 gratuity to the local swim team. We are currently attempting to determine the extent of this non- random recruiting problem. If we learn that the problem is widespri:ad, and that the data from these subjects is of questionable value, it may be appropriate to consider including a 13th city in the study. In order to facilitate more rapid demographic data analysis for these preliminary reports, we have initially focused on what is perceived to be the most important subject demographic and lifestyle data. We have compared some of that data with that of Cities No. 2 - 6, in order to place the Knoxville results in perspective. Results of the comparisons are provide below. CELL ASSIGNMENT In Figure 1 is compared the distribution of the Knoxville participants with that of the combined distributions of Cities 2 - 6, with regard to inclusion in smoking/non- smoking homes and workplaces. The relative individual cell populations for the first three cells (all tut NS/NS) are somewhat greater than those of the remaining cities. In the first six cities (except for F'ortland), no quota was placed on cell assignments resulting in a majority of participants falling into category 4 (no exposure at home or workplace). In Cities No. 5 & 6, cell populations in the first three cells was relatively low. Beginning with Daytona Beach (City #9) quotas were placed on cell assignments. The quotas should result in an approximately equal distribution among the four cells. SEX In Figure No. 2 is compared the participant sex distribution between Knoxville and those of the remaining cities. In Knoxville almost 80% of participants were female. For cities 2-6, females comprise almost 70% of the sample. RACE In Knoxville approximately 95% of the participants were white and about 4% were black (Figure No. 3). The general population of the Knoxville area (Knox county) is approximately 90% white and 9% black. For cities 2-6, whites comprised about 80% of the sample and l)lacks have comprised only about 2% of the sample. AGE The Knoxville participants (Figure No. 4) tended to be younger when compared to Cities 2-6, with 25% of the participants being less that 25 years old. Approximately 45% of participants were in the 25-39 age category. In the Knoxville area (Knox county), only about 25% of the residents fall into the 25-39 age category: For Cities 2-6 about 47% fall into this age category. 89272851 VBAfT- Af abts iir of a p.imk.ry atan rd srqittt to Mruisa sw ranctbnJ 3

EDUCATIONAL STATUS °-In Figure 5 is compared the relative educational statuses of the subject populations. Approximately 28% of the Knoxville participants had completed high school, compared to 18% for cities 2-6. Approximately 12% of the Knoxville participants had completed collE:ge as compared to 21% for cities 2-6. HOUSEHOLD INCOME In Figure No. 6 are compared the distributions of the household incomes. The household income for Knoxville participants is generally lower than for cities 2-6. Approa:imately 71% of the participating Knoxville households had incomes less that $40,000 per year as compared to only 50% for cities 2-6. SALIVARY COTININE Seven of the Knoxville subjects had mean salivary cotinine levels (average of all of the individual determinations for both "before" and "after" samples) greater than 10 ng/mL. This very high level places them in a category of "misclassified" non-smokers. That is, although they reported themselves to have not used any tobacco products or smoking cessation aids (transdermal nicotine patches or nicotine gum) for the past six months, there is no obvious way to explain their high salivary cotinine levels other than exposure to a concentrated source of nicotine. (Note that there can be dietary sources of nicotine, including potatoes, tomatoes, eggplant, etc., and that the dietary information provided by the subjects has not yet been thoroughly investigated. However, it seems unlikely that salivary levels of greater than 100 ng/mL could be accounted for through dietary exposure only.) The individuals may have deliberately lied to those administering the questionnaire, or they have mis-understood the questions. At any rate, we consider them to be mis-classified as non-s:nokers. Data summarizing some of the smoke exposure levels to which these individuals are exposed is summarized in Table 1. The levels of RSP and nicotine are sufficiently low to suggest that, if the individuals were smokers, they were not smoking around the personal sampling pump. Interestingly, three of the seven consider themselves to be never smokers. Because of the concerns regarding the reliability of the samples which these individuals may have collected, we have decided to exclude the smoke exposure data of these seven individuals until the data can be more closely examined. A9RRfT.- AI dstn a d 8 peWWry natwe mAs.bpct to mrsiW +nd tmnctanJ 4

Table 1. Smoke Constituent Levels in Homes of Mis-elassified Non-smokers (Salivary Cotinine ~ 10 ng/mL) PARTICIPAN SEX - EVER COTININE, RSP, NIC0TIN YEARS 10 C1G75 HOME ANY - OTHERS . , T!D SMOKED . END, UGlM"7 E ' SMOKED .-SMOKED ENVIROHMEN' SMKRS AT IN HOME' NGIML UGlM'] PER DAY T HOME? EXCLUSN ~ E OF SELF < 1007 MALE YES 270 60.6 1.70 6 30 SMOKING YES 2 1017 FEMAL NO 101 78.9 8.28 SMOKING YES 4 E 1024 MALE NO 268 31.6 0.64 NON-SMOKIN NO 4 G 1041 MALE YES 605 26.8 0.05 20 30 NON-SMOKIN NO 3 G 1047 FEMAL NO 165 26.0 7.78 SMOKING YES 4 E 1048 FEMAL YES 310 37.1 2.72 15 20 NON-SMOKIN NO 1 E G 1092 MALE YES 301 8.5 0.11 20 20 NON-SMOKIN NO 3 G ETS CONSTITUENT EXPOSURE In Table 2 are compared the median, low and high levels for smoke constituents measured in Knoxville in each of the four venues. Note that these are personal exposur5 levels in the venues, and do not represent the smoke exposure levels experienced by any one individual. Detailed data interpretation will be conducted at a later date. The data ranges are considerable for each category. However, it is clear that all of the median ETS constituent levels are higher in smoking environments than in non- smoking environments. This is graphically illustrated in Figures 7ELnd 8, which summarize the particle phase and vapor phase levels. For example, the median FPM and UVPM levels, indicative of combustion derived particulates, were a factor of 10 greater in srr oking homes than in non-smoking homes. 3-vinyl pyridine and nicotine were a factor of 7 and 10 greater in smoking workplaces than in non-smoking workplaces, respectively. In contrast, respirable susp.-nded particulate matter (RSP) was only about a factor of 2 greater in smoking environments. This is most likely due to the presence of other sources of RSP in most environments. With a few exceptions, most observed levels of ETS constituents were low (eg. RSP<100 µg/m'; nicotine < 5 µg/m). The levels of salivary cotinine observed for correctly classified subj,:cts are summarized in Table 3, and follow an expected pattern. That is, the median levels increase as the duration of exposure to V2ofT.• AI diti is of a parmnwr mtwe ind su*t to rerirsn sne can ctiwrl 5

Table 2. Summary of Smoke Exposure Data: Knoxville, TN (Corrected for blanks and Misclassified Subjects) Environment RSP (pg/M3) Solanesol (pg/M3) UVPM (pg/M') FPM (pg/M') 3-VP (pglM') Myosmine (,,g/M') Nicotine (pgM1') Non-smoking Home Median 18.000 0 1,225 0.690 0.052 0 0.050 Low 2.547 0 0.105 0.047 0 0 0.007 High 454.981 4.038 258.245 181.795 14.891 1.449 15.529 Smoking Home Median 32.832 0.125 13.328 9.348 1.105 0.139 1.413 Low 9.501 0 0.849 0.319 0.152 0.037 0.134 High 113.169 1.432 93.344 76.100 6.323 1.044 9.706 Non-smoking Workplace Median 19.658 0 1.148 U.4/4 u.u7s u ~.v5o Low 0 0 0 0 0 0 0 High 73.761 0.501 44.295 29.461 2.528 0.235 3.203 Smoking Workplace Median 34.805 0.033 5.850 3.330 0.476 0.065 0.530 Low 5.896 0 0.631 0.190 0 0 0.030 High 447.242 6.952 289.508 209.130 12.843 3.461 334.309 ,VnAfr. • w e.ra ir o/ a pdWVry ouwI ao wdyicr w mWoa aoe cmncrbnJ 6

ETS nicotine increases. In Knoxville, median levels increases in the order NS Home/NS Work<NS Home/S Work<S Home/NS Work<S Home/S Work. Each level inc:-eased by a factor of 3 over the _previous level. The range of levels is considerable, however, and more detailed analysis must be Performed to discern whether there is any correlation between smoke exposure and saliv cotinine " ~3' level. . Table 3. Mean Salivary Cotinine Levels in City No. 1 ENVIRONMENT Moan Salivary CoCinine, nglmL Media n Low High Smoking home/smoking workplace 3.71 0.48 5.56 Smoking home/non-smoking workplace 1.00 0 9.98 Non-smoking home/smoking workplace 0.30 0 5.48 i Non-smoking home/non-smoking workplace 0.1 -I 0 2.93 * Data excluded from those individuals thought to be mis-classified regarding smoking status (ie. mean cotinine level > 10 ng/mL) lDRAfT AI d~b is al a prmfminary natws inAs.~ict te ~erir~on snd tamtiw+J 7