Original Article | Published:

Estimation of environmental tobacco smoke exposure during pregnancy using a single question on household smokers versus serum cotinine

Journal of Exposure Analysis and Environmental Epidemiology volume 12, pages 286295 (2002) | Download Citation

Subjects

Abstract

Environmental tobacco smoke (ETS) exposure has been studied in relation to many diseases. The ability of a study to find an association between exposure and disease is, in part, determined by the accuracy of the exposure measure. This study examined how accurately one question, the number of smokers in the household, asked at birth, predicts ETS exposure in pregnant nonsmokers as assessed by serum cotinine. Blood specimens, drawn at 15–19 weeks gestation, from 783 women who participated in a prenatal screening program in California in 1992 were analyzed for cotinine. Serum cotinine was significantly correlated with the number of smokers in the household (r=0.35, P<0.001, geometric mean cotinine (nanograms per milliliter) for 0 smokers=0.06, 1 smoker=0.18, 2 or more smokers=0.29). Using multiple regression, the number of smokers in the household accounted for 11% of the variation in serum cotinine. Cotinine concentrations were twice as high in women living with one or more smokers compared to women not living with a smoker, when reported exposure (0 or >0h) at home, work and other places was similar. Thus, the number of household smokers can account for a statistically significant amount of variation in serum cotinine and omission of this information would result in an underestimation of ETS exposure. Although use of this question alone does not provide an adequate estimation of ETS exposure as determined by serum cotinine, the results of this study indicate that this question is an important component of assessing ETS exposure.

Introduction

For more than two decades, exposure to environmental tobacco smoke (ETS) has been reported to be associated with numerous adverse health effects in nonsmokers (US DHHS, 1980; Spitzer et al., 1990). Nonsmokers exposed to ETS are at increased risk for certain cancers (US EPA, 1992), cardiovascular disease (Glantz and Parmley, 1995; Taylor et al., 1992), and adverse reproductive and developmental outcomes (OEHHA, 1997; US DHHS, 2001). How precisely ETS exposure is measured can strongly affect a study's ability to correctly quantify the relation between exposure to ETS and the risk of disease or adverse outcome. In such a study, misclassifying women with ETS exposure as unexposed would bias results toward the null hypothesis.

Although exposure of the pregnant woman to ETS during pregnancy has been associated with lower birth weight (Haddow et al., 1987; Mathai et al., 1992; Saito, 1991; Mainous and Hueston, 1994; Martinez et al., 1994; Roquer et al., 1994), in many studies the effect seems to be modest (Martin and Bracken, 1986; Chen et al., 1989; Zhang and Ratcliffe, 1993; Eskenazi et al., 1995; Rebagliato et al., 1995a). Some of these studies have used questionnaire data to determine ETS exposure (Martin and Bracken, 1986; Chen et al., 1989; Mathai et al., 1992; Fortier et al., 1994; Roquer et al., 1994; Dejin-Karlsson et al., 1998), with five studies relying solely on information concerning household members who smoke, one as recently as 1997 (Brooke et al., 1989; Chen et al., 1989; Mathai et al., 1992; Saito, 1991; Ahluwalia et al., 1997). Other studies examining the association between ETS exposure and low birth weight have measured exposure using a biomarker such as cotinine (Eskenazi et al., 1995; Haddow et al., 1988; Peacock et al., 1988). Studies that determine exposure using a biomarker are more difficult to conduct and more expensive than questionnaire data. Studies have examined the agreement between self-reported ETS exposure and biochemical markers of ETS exposure (Coultas et al., 1988; Coultas et al., 1989; Coghlin et al., 1989; Haley et al., 1989; Riboli et al., 1990; Becher et al., 1992; Emmons et al., 1994; Kemmeren et al., 1994), and a few have examined this association in pregnant women (O'Connor et al., 1995; Rebagliato et al., 1995b). However, the analytic methods used to determine biochemical markers of ETS exposure in many of these studies were not as sensitive as can be obtained currently. Recent advances in laboratory methods enable detection of the very small amounts of cotinine in the blood resulting from low-level ETS exposure (Bernert et al., 1997). In studies that used less sensitive cotinine assays, women exposed to low levels of ETS would have been misclassified as nonexposed. Cotinine, a metabolite of nicotine, is considered the most specific and most sensitive biomarker for ETS exposure (Jarvis et al., 1987; Benowitz, 1996), and has been used to validate other measures of exposure to ETS.

Past studies have used a single question to evaluate ETS exposure, and future assessment of ETS exposure may also use a single question, such as might appear on a birth certificate. Thus, it would be very important to know how well such a question reflects ETS exposure to assess exposure across groups, to predict exposure prevalence, or to characterize individual exposure. Therefore, the purpose of this paper is to determine how accurately a single question, concerning the number of smokers in the household, can predict ETS exposure in mid-pregnancy as assessed by serum cotinine concentration.

Materials

Study Population

The overall study population included pregnant women from a four-county region of central California (Fresno, Kern, Kings, and Tulare), who enrolled in the State's Maternal Serum Alpha-Fetoprotein (MSAFP) prenatal screening program in April 1992 (N=1560). Prior to participation, women read and signed an informed consent. The study was approved by the California Health and Human Services Agency Committee for the protection of human subjects. Approximately 60% of women delivering live births enrolled in the screening program in 1992. As part of the screening program, a woman provided a blood specimen between the 15th and 19th weeks of gestation. Blood specimens were collected in serum separator tubes and sent via regular mail to a regional laboratory where they were analyzed for MSAFP. The remainder of each specimen was stored for up to 4 years at −20°C until it was sent to the Centers for Disease Control and Prevention (CDC) for cotinine analysis.

At the time of birth registration, a one-page environmental health questionnaire (supplemental to the birth questionnaire) was administered in either English or Spanish in the hospital by the birth recorder. The 22-item questionnaire included two questions on ETS exposure, including the question of interest in the present study: “Excluding yourself, how many members of your household smoked cigarettes during the fourth and fifth months of pregnancy?” A second ETS question asked for the number of hours a woman was exposed to ETS at each of the following locations: home, work, or other places, in the fourth and fifth months of pregnancy. While information from the second ETS question is included in selected analyses in this paper, a more detailed analysis is presented elsewhere (DeLorenze et al., 2002). Prenatal screening program records and specimens were linked to data from live birth certificates and environmental health questionnaires using maternal identifiers, such as names and date of birth, in a probabilistic record linkage software package (Jaro, 1995).

Of the 1560 eligible women, some were excluded from the analysis if: there was no cotinine sample (n=229); there was no match with vital statistics records (n=51); the environmental health questionnaire was not given to the woman (n=66), or the environmental health questionnaire was not completed by the woman (n=176). Of the remaining 1038 women, 61 women (6%) did not respond to the ETS question of interest; these women were excluded from further analysis. In attempting to insure that the final sample of women included only nonsmokers, an additional 194 women were excluded if they indicated on the prenatal screening record or the environmental health questionnaire that they smoked in the 3 months before pregnancy or at any time during the pregnancy, or if they had a serum cotinine value above 10 ng/ml (n=16, range of cotinine 12–110 ng/ml). A cutoff of 10 ng/ml was used to exclude even occasional smokers. Thus, 783 women were included in the analysis.

The 61 women (6%) who failed to complete the question about the number of smokers in the household were significantly different from the responders. A statistically significant greater percentage of these women: had fewer years of schooling; were of Hispanic or other non-White non-Black ethnic origin; completed the questionnaire in Spanish; spoke a language other than English at home; and had a government-funded program as their principal source of payment for prenatal services. After excluding women based on the criteria noted above, the remaining 56 women had a geometric mean serum cotinine concentration of 0.08 ng/ml (95% CI=0.05, 0.12), which was similar to the final sample.

Serum Cotinine Analysis

The serum cotinine concentration was measured at CDC using a sensitive isotope-dilution high-performance liquid chromatographic/atmospheric pressure ionization tandem mass spectrometric (LC/MS/MS) procedure (Bernert et al., 1997). Samples were preclassified by a screening cotinine immunoassay and subdivided into nominal “high” or “low” cotinine groups. The samples were then analyzed by LC/MS/MS in runs of 50 (either all high or all low), which included two water blanks and two quality control (QC) pool sera. The low run QC pool averaged 1.77±0.094 ng/ml (CV=5.3%) throughout these analyses. These samples were analyzed over a period of several months in two different calibration groups. Limits of detection (LOD) were calculated by the limiting standard deviation method (Taylor, 1987) for these two groups, and remained below the nominal method detection limit of 0.050 ng/ml in each case. Serum cotinine concentration was log normally distributed (determined using a Box–Cox test; Box and Cox, 1964).

Approximately one-third of the specimens had cotinine values below the LOD. To preserve the integrity of the log normal distribution (i.e., to avoid placing an unnaturally large proportion of the study population at a single value at the lower end of the distribution), actual cotinine values below the LOD were used instead of assigning the LOD or 0.5*LOD to these specimens. In addition, to retain 29 specimens with cotinine values of zero in the analyses after log transformation, the value of the lowest measured cotinine sample (0.001 ng/ml) was assigned to these specimens. The statistical analyses (described below) were rerun in three other ways (assigning 0.5*LOD to specimens with values below the LOD, assigning the LOD, and excluding such specimens) and the results were not meaningfully altered.

Statistical Analysis

All statistical analyses were conducted using Stata (Stata, version 6.0). The distribution of the number of smokers in the household and serum cotinine concentration was examined using univariate analyses (Kruskal–Wallis test). These two distributions were then compared using Spearman rank order correlation and scatter plots. Multiple linear regression was used to examine the factors contributing to the variance in serum cotinine concentrations, specifically the number of smokers in the household. The distribution of serum cotinine concentrations was skewed to the left and therefore all regression analyses were conducted on log-transformed values. The raw data of the number of smokers in the household included categories 0–5; however, there were too few observations in the highest three categories so the number of smokers was classified as 0, 1, or 2 or more smokers. The number of smokers in the household was examined in two ways in the multiple regression models, as a categorical (0, 1, 2+), or dichotomous variable (0, 1+). The influence of hours of ETS exposure at home, work, and in other places was examined in this study coding each ETS location variable as zero or greater than zero hours of exposure. Covariates considered in the model included maternal age, education, ethnicity, marital status and payment source for prenatal care, and parity derived from the birth certificate. Variables derived from the environmental health questionnaire, such as whether the women worked outside the home during pregnancy, language spoken at home, language used to complete the environmental health questionnaire, caffeine consumption, and tea consumption were also examined in the model. Marital status was inferred from information supplied on the birth certificate concerning the father. Racial/ethnic groups were identified as non-Hispanic White, Black, USA-born Hispanic, Mexican-born Hispanic, and Other. The Hispanic ethnic groups were restricted to women who indicated that their Hispanic origin was Mexican, Mexican American or Chicano. Women of other Hispanic origin such as Puerto Rican, Central or South American were classified as Other. Women within the Hispanic group were classified as USA-born or Mexican-born; however, since no significant differences were observed between these groups in further analyses the two groups were subsequently combined as one Hispanic group. Interactions between number of smokers in the household and the covariates mentioned above were also examined; however, none was detected. Logistic regression was used to identify factors related to women who had reported no smokers in the household yet had unexpectedly high serum cotinine concentrations and, conversely, to identify factors related to women who had reported one smoker in the household yet had unexpectedly low serum cotinine concentrations. The linktest was used to determine that the model was specified correctly (Tukey, 1949; Pregibon, 1979; Pregibon, 1980).

Results

Of the 783 respondents, 79% reported no smokers in the household, 17% reported one smoker in the household, and 4% reported 2 or more smokers in the household. The distribution of the study population according to selected demographic characteristics, the geometric mean cotinine concentrations, and the number of smokers in the household is presented in Table 1. A greater percentage of younger nonsmoking women were exposed to one or more smokers in the home compared with older nonsmokers. The percentage of women exposed to smokers in the household also differed (P<0.05) by education, marital status, type of payment for prenatal services, and whether the woman worked outside the home during pregnancy. Geometric mean serum cotinine concentration of all women was 0.07 ng/ml (95% confidence interval [CI] 0.06, 0.08 ng/ml) (median serum cotinine=0.07 ng/ml, range of 0–7.96 ng/ml, mean serum cotinine concentration=0.20 ng/ml, SD=0.48 ng/ml.). The geometric mean cotinine, presented in Table 1, varied (P<0.05) within the following subgroups: education, ethnicity, marital status, type of payment for prenatal services, and whether the woman worked outside the home during pregnancy. The distributions of serum cotinine concentrations were similar to the distributions of reported ETS exposure as measured by the number of smokers in the home.

Table 1: Selected demographic characteristics by the number of smokers in the household of pregnant, nonsmoking, prenatal screening enrollees.

As shown in Table 2, geometric mean serum cotinine concentration in nonsmokers increased with increasing number of smokers in the household. The range of cotinine values illustrates the large degree of variation among the categories of smokers in the home. Serum cotinine concentration was significantly correlated with the number of smokers in the home (Spearman correlation=0.35, P<0.001). Figure 1 illustrates the normalized distribution of serum cotinine concentration on a log scale for women living with no smokers, one smoker, or two or more smokers. Although there is considerable overlap among the distributions, there is clearly a shift to the right, indicating higher serum cotinine values, as the number of smokers in the home increases.

Table 2: Geometric mean serum cotinine concentrations for number of smokers in the household of pregnant, nonsmoking, prenatal screening enrollees
Figure 1
Figure 1

Serum cotinine (ng/mL) by number of smokers in the household. Normalized distribution of serum cotinine (ng/mL) by number of smokers in the household (0, 1, 2+) for pregnant, nonsmoking, prenatal screening enrollees.

Table 3 shows the geometric mean serum cotinine concentrations for women who either lived with a smoker or did not (according to the first ETS question), and were either exposed to ETS in the home or were not, i.e., having greater than zero hours or zero hours of exposure (according to the second ETS question). Women who reported living with a smoker and were not exposed to ETS in the home had twice the cotinine concentrations (0.10 ng/ml) than women who reported not living with a smoker and were not exposed to ETS in the home (0.05 ng/ml, P<0.001, Wilcoxon rank sum test). Exposure to ETS in places other than the home did not explain this difference because serum cotinine concentrations for women who were not exposed to ETS anywhere, including at home, work, or in other places, were higher in those women who lived with a smoker than in women who did not live with a smoker (0.08 vs. 0.04 ng/ml, respectively, P<0.005). In fact, geometric mean serum cotinine concentrations were approximately double for women living with a smoker compared with women not living with a smoker, for each ETS exposure situation.

Table 3: Serum cotinine concentrations by smokers in the home and ETS exposure at various locations for pregnant, nonsmoking, prenatal screening enrollees

Using regression analysis, the number of smokers in the household accounted for 11% of the variation in log serum cotinine concentrations (Table 4, Model 1). Responses to the second ETS question, i.e., whether a woman was exposed to tobacco smoke at home, at work, or in other places, were considered in the regression models 2 and 4. Exposure at work was not a significant factor. Exposure to ETS at home and exposure in other places were statistically significant variables in the model and increased the amount of explained variation (adjusted R2=16%, Model 2). The results were similar whether the number of smokers in the household was entered into the regression equations as either a categorical or dichotomous variable. Four covariates (maternal age, marital status, intended source of payment for prenatal services, maternal ethnicity being “Hispanic”) were found to be statistically significant when added to Models 1 and 2 (shown in Models 3 and 4, respectively). The amount of variation explained by the model, following the addition of these covariates, increased by only 5% to explain a total of 21% (Model 2 vs. Model 4), while the regression coefficient for the number of smokers in the household decreased slightly (β=0.45, Model 2 vs. β=0.34, Model 4).

Table 4: Multiple linear regression models of serum cotinine and ETS exposure in pregnant, nonsmoking, prenatal screening enrollees

Of the women who responded that they do not live with a smoker or that they live with one smoker, only 2% and 53%, respectively, responded that they were exposed to ETS in the home. As shown in Table 2, there was considerable variation in the cotinine values for women who reportedly did not live with a smoker or who lived with only one smoker. Given this wide range, we wanted to determine which variables predicted who reported not living with a smoker but whose serum cotinine concentrations were very high, and who reported living with a smoker but whose serum cotinine concentrations were very low. To examine the discordance in women who reported not living with a smoker but who had higher levels, we dichotomized the group by serum cotinine concentration into the upper quartile (0.131 to 5.17 ng/ml) versus the lower three quartiles (0.001 to 0.130 ng/ml). Similarly, serum cotinine concentrations for women who reported living with one smoker were dichotomized into the lowest quartile (0.001–0.072 ng/ml) versus the upper three quartiles (0.073–7.96 ng/ml). Table 5 shows results of the logistic regression analysis of serum cotinine level, (upper quartile=1 and lower quartiles=0 for zero smokers in the home; lowest quartile=1 and the higher three quartiles=0 for one smoker in the home) in relation to other ETS exposure measures as well as demographic characteristics. Among women who reported not living with a smoker, the odds of women being in the upper quartile versus the lower three quartiles of serum cotinine concentrations were significantly greater in women who reported exposure to ETS at home and in other places; who were unmarried; of “Other” ethnicity; and who expected to pay for prenatal services through a government program. Among women who reported living with a smoker, the odds of a women being in the lowest quartile versus the upper three quartiles of serum cotinine concentrations increased with increasing maternal age and were greater in women who reported not being exposed to ETS at home or in other places. Consistent with the previous observation for women not living with a smoker, no exposure to ETS at home or in other places was significantly associated with lower serum cotinine concentration. Conversely, ethnicity, marital status, and type of payment for prenatal services were not significantly associated with classification in the lowest quartile of serum cotinine concentration, while increasing maternal age was significant.

Table 5: Logistic regression of serum cotinine concentrations, in relation to ETS exposures and demographic variables in pregnant, nonsmoking, prenatal screening enrollees

Discussion

The present study found that a question on the number of smokers in the household was a significant factor in determining exposure to ETS as assessed by serum cotinine concentrations among nonsmoking pregnant women in California. Whether or not a woman reported having smokers in the household explained 11% of the variation in blood cotinine concentrations. Our findings are in agreement with other studies (Riboli et al., 1990; Emmons et al., 1994; Pirkle et al., 1996). Emmons et al. (1994) also reported that the number of smokers in the household accounted for 11% of the variation in salivary cotinine concentrations. This is in spite of the fact that the populations differed considerably with only 63% of their population women, not necessarily pregnant, who were older, and had more years of education. However, in a study by O'Connor et al. (1995), which included only pregnant women, and found a significant correlation between the number of smokers in the home and air nicotine (r=0.35), the correlation with urinary cotinine (r=−0.02) was not significant. This is in contrast to the correlation observed in the current study between the number of smokers and serum cotinine (r=0.35). This discrepancy between the results of the present study and those of O'Connor's study may be due to a higher proportion of study subjects falling below the detection level of the assay in the latter study, 48%, versus 36% in the present study. These two studies may not be directly comparable given that cotinine in urine is approximately six times more concentrated than in serum and may have greater individual variation (Benowitz, 1996).

The present study showed that the use of a single question on household smokers would provide a useful measure of the prevalence of ETS exposure in pregnant women at a population level. It also showed that this question could rank subgroups of the pregnant population according to their relative exposure to ETS in a similar manner as observed using serum cotinine concentrations (Table 1). The response to this question, however, may not adequately reflect an individual pregnant woman's ETS exposure (Table 2). This misclassification from usage of questionnaire data may have adversely affected the ability of previous studies to detect potentially harmful effects of ETS exposure on outcomes such as birth weight by likely lowering the magnitude of any true association.

Information about the number of smokers in the household, reportedly the site of greatest exposure to ETS (Haley et al., 1989; Emmons et al., 1994) consistently explained a significant amount of variation in serum cotinine concentrations regardless of additional information on ETS exposure provided by the second question — exposure at home, work or other places. Serum cotinine concentrations were approximately twice as high in women living with a smoker compared with women who were not living with a smoker in each of the four ETS-exposure settings. These findings greatly emphasize the importance of this question in assessing exposure to ETS. In the regression analysis, the number of smokers in the home explained nearly twice as much variance as the six covariates combined; however, the classifications of ETS exposure in the home and in other places were more significant than the number of smokers in the household. Although the number of smokers in the household may not be the most important question to ask, the findings of the present study strongly support the premise that it is, however, an important factor in assessing ETS exposure. Therefore, the results of the present study suggest that disregarding the question concerning the number of smokers in the household and asking solely about specific hours of exposure would result in an underestimate of ETS exposure, as assessed by the serum cotinine concentrations.

Higher serum cotinine concentrations in women who lived with a smoker compared with women who did not live with a smoker may be a result of exposure to ETS in other places that was not adequately measured in this study. This may, however, also be a result of exposure to nicotine emitted into the air from a smoker's clothes or hair even when the smoker has not smoked in the home (Nelson et al., 1991; Benowitz, 1996). Nicotine has also been measured in the air 3 days after smoking occurred in a room (Nelson et al., 1991). Nelson et al. (1991) further showed subjects with increased urinary cotinine concentrations after exposure to air in a test chamber where nicotine was present from the “desorption of previously adsorbed nicotine” from the walls of the chamber. Nicotine, even at lower concentrations and separate from the particulate phase of ETS, may be an important factor in decreased uteroplacental perfusion and subsequent lower infant birth weight (Lambers and Clark, 1996).

In attempting to determine factors related to the large variation in serum cotinine concentrations for women who had reported not living with a smoker, exposure to ETS in other places and in the home were shown to be significantly associated with higher serum cotinine. This further emphasizes the need for a question asking about hours of exposure in various locations. In addition, being single, being of non-White, non-Hispanic, and non-Black ethnicity, and intending to pay for prenatal services through a government program were also found to be significant factors in the likelihood of being in the highest quartile of serum cotinine concentrations. Significant predictors of lower serum cotinine in women who reported living with one smoker included reporting no ETS exposure in other places and at home, as well as increasing maternal age. Some of this misclassification may be explained by not capturing all ETS exposures [i.e., by not asking the right question(s)] and possibly by the subjects not responding accurately. It may be that the measurement of ETS exposure from women with the above-mentioned characteristics might not be as accurate.

The large amount of unexplained variation in serum cotinine concentrations may be the result of one or more of the following factors. Firstly, no information was collected on whether the smoker actually smoked inside the house. The potential for exposure misclassification is illustrated by the large percentage of women who reported living with a smoker but reported not having any exposure to ETS in the home (47% of women living with one smoker, 15% of women living with two or more smokers). Whether or not the woman was exposed at home was the most significant variable in the regression analysis. As shown in other studies (DeLorenze et al., 2002; Rebagliato et al., 1995b), including a quantitative estimate of ETS exposure in various locations explains a significant amount of variation in serum cotinine concentrations. Secondly, the range of response to the ETS exposure question was small in this sample of women, thus limiting the ability to explain the variation in serum cotinine. Only 21% of nonsmoking women reported living with one or more smokers, with only 4% of those living with two or more smokers, and only 28% reported being exposed to ETS anywhere. Thirdly, the size of the home and the amount of ventilation in the home may be important in determining ETS exposure (Cummings et al., 1990; Riboli et al., 1990; Pirkle et al., 1996; Henschen et al., 1997). This information was not included in the present study.

In acknowledging the limitations of this study it must be recognized that a one-time measure of cotinine at 15–19 weeks of gestation is not a “gold standard” for the assessment of ETS exposure. Two concerns of using cotinine as a biomarker of ETS are (1) that it does not provide a measure of long-term exposure since it has a half-life of approximately 20 h (Benowitz, 1999), and (2) that there are interindividual differences in the rates of nicotine and cotinine metabolism (Benowitz and Jacob, 1994; Perez-Stable et al., 1998). Another limitation of the current study is that the ETS questions about exposure at 4–5 months of pregnancy were asked months later at the time of birth. It is likely, however, that the woman's recall for the question regarding the number of smokers in the home would not have been as adversely affected as recall for questions asking about hours of exposure in specific locations. Lastly, in California, exposure to ETS in the workplace as well as in buildings is limited due to strict smoking laws. Therefore, ETS exposure in this population of California women may be unique in comparison to women in other states.

The strengths of this study include the ethnically diverse, population-based sample of women, the use of a cotinine assay with a very low detection limit and the high response rate to the ETS question. This sample of women includes Whites, Blacks, and Hispanics, a group not previously studied with respect to exposure to ETS. The extremely sensitive cotinine assay enabled the detection of ETS exposure at levels far below most studies. The proportion of women in this sample with cotinine values below the limit of detection was low in comparison to other studies. Thus, this study was able to detect ETS exposure in many women that other studies, with less sensitive assays, would have classified as unexposed. The question regarding the number of smokers in the home is relatively easy to answer and had a high response rate (94%), higher than the ETS question concerning hours of exposure at home, work, or other places (DeLorenze et al., 2002), and is not likely to have been subject to misreporting several months later.

These results suggest that a single question concerning the number of smokers in the household can account for a statistically significant amount of variation in serum cotinine concentrations, although the use of this question alone does not provide an adequate estimation of ETS exposure as compared with serum cotinine. Nonetheless, there is strong indication that information regarding the number of smokers in the home is an important component of assessing exposure to ETS and omission of this information would result in an underestimation of ETS exposure.

References

  1. Exposure to environmental tobacco smoke and birth outcome: increased effects on pregnant women aged 30 years or older, Am J Epidemiol (1997) 146: 42–47

  2. Passive smoking in Germany and Poland: comparison of exposure levels, sources of exposure, validity, and perception, Epidemiology (1992) 3: 509–514

  3. Cotinine as a biomarker of environmental tobacco smoke exposure, Epidemiol Rev (1996) 18(2): 188–204

  4. Biomarkers of environmental tobacco smoke exposure, Environ Health Perspect (1999) 107(Suppl 2): 349–355

  5. Metabolism of nicotine to cotinine studies by a dual stable isotope method, Clin Pharmacol Ther (1994) 56: 483–493

  6. Development and validation of sensitive method for determination of serum cotinine in smokers and nonsmokers by liquid chromatography/atmospheric pressure ionization tandem mass spectrometry, Clin Chem (1997) 43(12): 2281–2291

  7. An analysis of transformations, J R Stat Soc (1964) B-26: 211–243 discussion 244–252

  8. Effects on birthweight of smoking, alcohol, caffeine, socioeconomic factors and psychosocial stress, Br Med J (1989) 298(6676): 796–801

  9. Passive smoking and low birth weight [letter] Lancet 2 (1989) 8653: 54–55

  10. Development of epidemiologic tools for measuring environmental tobacco smoke exposure, Am J Epidemiol (1989) 130(4): 696–704

  11. Discrepancies between self-reported and validated cigarette smoking in a community survey of New Mexico Hispanics, Am Rev Respir Dis (1988) 137: 810–814

  12. Questionnaire assessment of lifetime and recent exposure to environmental tobacco smoke, Am J Epidemiol (1989) 130(2): 338–347

  13. Measurement of current exposure to environmental tobacco smoke, Arch Environ Health (1990) 45(2): 74–79

  14. Does passive smoking in early pregnancy increase the risk of small-for-gestational-age infants?, Am J Public Health (1998) 88(10): 1523–1527

  15. . Environ Res (2002) (in press)

  16. An evaluation of the relationship between self-report and biochemical measures of environmental tobacco smoke exposure, Prev Med (1994) 23(1): 35–39

  17. Passive and active maternal smoking as measured by serum cotinine: the effect on birthweight, Am J Public Health (1995) 85: 395–398

  18. Passive smoking during pregnancy and the risk of delivering a small-for-gestational-age infant, Am J Epidemiol (1994) 139(3): 294–301

  19. Passive smoking and heart disease. Mechanisms and risk, JAMA (1995) 273: 1047–1053

  20. Cigarette consumption and serum cotinine in relation to birthweight, Br J Obstet Gynaecol (1987) 94: 678–681

  21. Second-trimester serum cotinine levels in nonsmokers in relation to birth weight, Am J Obstet Gynecol (1988) 159(2): 481–484

  22. Biochemical validation of self-reported exposure to environmental tobacco smoke, Environ Res (1989) 49: 127–135

  23. The internal dose of passive smoking at home depends on the size of the dwelling, Environ Res (1997) 72(1): 65–71

  24. Probabilistic linkage of large public health data files, Stat Med (1995) 14: 491–498

  25. Comparison of tests used to distinguish smokers from nonsmokers, Am J Public Health (1987) 77(11): 1435–1438

  26. Plasma cotinine: stability in smokers and validation of self-reported smoke exposure in nonsmokers, Environ Res (1994) 66(2): 235–243

  27. The maternal and fetal physiologic effects of nicotine, Semin Perinatol (1996) 20(2): 115–126

  28. Passive smoke and low birth weight, Arch Fam Med (1994) 3: 875–878

  29. Association of low birth weight with passive smoke exposure in pregnancy, Am J Epidemiol (1986) 124: 633–642

  30. The effect of paternal smoking on the birthweight of newborns whose mothers did not smoke, Am J Public Health (1994) 84(9): 1489–1491

  31. Passive maternal smoking and birthweight in a South Indian population, Br J Obstet Gynaecol (1992) 99: 342–343

  32. Where there's smoke…? Biases in the use of nicotine and cotinine as environmental tobacco smoke biomarkers In Proceedings of the 1991 EPA/A&WMA International Symposium: Measurement of Toxic and Related Air Pollutants. Vol. 1 Air and Waste Management Association, Pittsburgh, PA 1991 449–454

  33. Measurement of exposure to environmental tobacco smoke in pregnant women, Am J Epidemiol (1995) 142: 1315–1321

  34. Office of Environmental Health Hazard Assessment (OEHHA) . Health Effects of Exposure to Environmental Tobacco Smoke. California Environmental Protection Agency, Sacramento, CA 1997

  35. Maternal cotinine level during pregnancy and birthweight for gestational age, Int J Epidemiol (1988) 27(4): 647–656

  36. Nicotine metabolism and intake in black and white smokers, JAMA (1998) 280(2): 152–156

  37. Exposure of the US population to environmental tobacco smoke, JAMA (1996) 275(16): 1233–1240

  38. Data Analytic Methods for Generalized Linear Models PhD dissertation University of Toronto 1979

  39. Goodness of link tests for generalized linear models, Appl Stat (1980) 29: 15–24

  40. Exposure to environmental tobacco smoke in nonsmoking pregnant women in relation to birth weight, Am J Epidemiol (1995a) 142: 531–537

  41. Assessment of exposure to environmental tobacco smoke in nonsmoking pregnant women in different environments of daily living, Am J Epidemiol (1995b) 142: 525–530

  42. Exposure of nonsmoking women to environmental tobacco smoke: a 10-country collaborative study, Cancer Causes Control (1990) 1: 243–252

  43. Influence on fetal growth of exposure to tobacco smoke during pregnancy, Acta Paediatr (1994) 84: 118–121

  44. The smoking habits of pregnant women and their husbands, and the effect on their infants, Nippon Koshu Eisei Zasshi (1991) 38: 124–131

  45. Links between passive smoking and disease: a best-evidence synthesis. A report of the working group on passive smoking, Clin Invest Med (1990) 13: 17–42

  46. Environmental tobacco smoke and cardiovascular disease. A position paper from the council on cardiopulmonary and critical care, American Heart Association, Circulation (1992) 86: 699–702

  47. Principles of measurement In Quality Assurance of Chemical Measurements. Lewis Publishers, Chelsea, MI 1987 75–93

  48. One degree of freedom for non-additivity, Biometrics (1949) 5: 232–242

  49. U.S. Department of Health and Human Services (USDHHS) The health consequences of smoking for women: a report of the Surgeon General US DHHS publication no. 410-889/12840 U.S. Department of Health and Human Services, Rockville, MD 1980

  50. U.S. Department of Health and Human Services (USDHHS). Women and smoking: a Report of the Surgeon General. USDHHS, CDC, National Center for Chronic Disease Prevention and Health Promotion, Office of Smoking and Health: Atlanta, GA 2001

  51. U.S. Environmental Protection Agency (US EPA) Respiratory health effects of passive smoking: lung cancer and other disorders EPA publication no. 600/6-90/006F Off. Health Environ. Assess., US EPA, Washington, DC 1992

  52. Paternal smoking and birthweight in Shanghai, Am J Public Health (1993) 83(2): 207–210

Download references

Acknowledgements

This research was supported by the California Tobacco-Related Disease Research Program (Grant No. 6RT-0385). The data collection was aided by March of Dimes Birth Defects Foundation Grant Nos. 15-FY92-0078 and 15-FY93-0662. The authors thank Lynn Goldman, George Cummingham, Robert Haas, Richard Kreutzer, Enid Satariano, Steve Graham, Cindy Evangelista, and Betsy Noth.

Author information

Affiliations

  1. Public Health Institute, Berkeley, California, USA

    • FARLA LYNN KAUFMAN
    •  & GERALD NICHOLAS DELORENZE
  2. California Department of Health Services, Genetic Disease Branch, Berkeley, California, USA

    • MARTIN KHARRAZI
  3. University of California at Berkeley, Berkeley, California, USA

    • BRENDA ESKENAZI
  4. Centers for Disease Control and Prevention, Atlanta, Georgia, USA

    • JOHN THOMAS BERNERT

Authors

  1. Search for FARLA LYNN KAUFMAN in:

  2. Search for MARTIN KHARRAZI in:

  3. Search for GERALD NICHOLAS DELORENZE in:

  4. Search for BRENDA ESKENAZI in:

  5. Search for JOHN THOMAS BERNERT in:

Corresponding author

Correspondence to FARLA LYNN KAUFMAN.

About this article

Publication history

Received

Published

DOI

https://doi.org/10.1038/sj.jea.7500224

Further reading