774 



Coultas, Stidley, and Samet FTC Tar and Nicotine Yields and Markers of Exposure to Tobacco Smoke 



437 



3 o« - a m 



19 00 - 19 t 



Figure 1. 

 to 1985 



Tar Categories (mg) 

 Distributions of FTC cigarette tar yields m a general population sample of Hispanics, New Mexico, 1984 



nine and end-expired carbon monoxide levels, r ^ 0.64 (N =: 264. 

 p < 0.0001). For other measures of exposure (table 2). moderate 

 levels of correlation were observed between current number of 

 cigarettes smoked per day and the biologic markers salivary coti- 

 nine and end-expired carbon monoxide (figures 2 and 3). There 

 was a weak association between the tar level and the salivary coti- 

 nine level. 



We developed multiple linear regression models to character- 

 ize further the predictors of salivary cotinine and end-expired car- 

 bon monoxide (tables 3 and 4) Several models were examined 

 using available independent variables as predictors of salivary 

 cotinine and end-expired cart)on monoxide levels (see Methods). 

 Current number of cigarettes per day and time since smoking the 

 last cigarette were statistically significant predictors in all models 

 (tables 3 and 4). The current number of cigarettes smoked was 

 consistently the most important predictor. Addition of FTC yields 



to the models explained only a small proportion of the vanability 

 in cotinine and carbon monoxide levels. Models that included in- 

 halation patterns as categorical variables (0 ^ into top of lungs or 

 deeply into lungs; 1 = into mouth only or into back of throat) 

 showed lower levels of cotinine (B = -11.14 ng/ml.p » 0.65) and 

 carbon monoxide levels (B = -2.39 ppm. p - 0.16) associated 

 with reports of inhaling only into the mouth or throat, but the coeffi- 

 cients were not statistically significant. 



We also used multiple linear regression to examine predictors 

 of the current number of cigarettes smoked per day. with sex, age. 

 ethnicity, and FTC tar and nicotine yields as independent vari- 

 ables. Significant predictors were sex (on average, males smoked 

 4.1 cigarettes more than females, p = 0.002). age (on average, 

 subjects less than 50 yr of age smoked 32 cigarettes more than 

 those 50 yr of age and older, p = 0.02). and ethnicity (on average. 

 Hispanics smoked 10.9 fewer cigarettes than non-Hispanics, p < 

 0.0001). FTC tar and nicotine yields were not significant. 



TABL£ 2 



SPEARMAN CORBEUVTION COEFFICIENTS BETWEEN VARIOUS 



MEASURES OF CIGARETTE SMOKE EXPOSURE AMONG 



ACTIVE SMOKERS IN NEW MEXICO. 19S4 TO 1985 



N 1 



Salivary cotinina. ng/ml 



Current number of cigarettes smolted/day 



FTC lar. mg 



FTC nicotine, mg 



FTC CO. ppm 

 End-tidal carten morTOKidfl. ppm 



Current number of cigarettes smolred/day 



FTC tar. mg 



FTC nicotine, mg 



FTC CO. ppm 



- p - 00001 

 t p - O03 



DISCUSSION 



In this largely Hispanic community sample, some indices of ex- 

 posure to tobacco smoke from active smoking varied widely be- 

 tween the sexes and ethnic groups. We found a broad range of 

 values lor numbers of cigarettes, salivary cotinine levels, and end- 

 expired carlwn monoxide levels. However, types of cigarettes 

 smoked were more narrowly distributed. Most subjects (90%) 

 smoked filter cigarettes, and the majority smoked cigarettes of 

 medium or high tar content (> 10 mg/cigarene). For all smokers, 

 regardless of cigarette type, there was a moderate associsliio 

 between the current number of cigarettes smoked per day and 

 levels of the biologic markers salivary cotinine and end-expired 

 carbon monoxide (table 2) In addition, our findings provide fur- 

 ther documentatkHi that FTC figures on cigarette yieWs of tar. nk» 

 tine, and cartwn monoxide are not strongly predictive of exposure 



