Exposure to secondhand smoke and associated factors among non-smoking pregnant women with smoking husbands in Sichuan province, China



    1. Department of Health Economics, Huaxi School of Public Health, University of Sichuan, Chengdu, China
    2. Department of Health Economics, School of Public Health and Administration, Chengdu University of Traditional Chinese Medicine, China
    Search for more papers by this author

    1. Department of Internal Medicine, University of California, Davis, USA
    Search for more papers by this author

    Corresponding author
    1. Department of Health Economics, Huaxi School of Public Health, University of Sichuan, Chengdu, China
      Zhengzhong Mao, Department of Health Economics, Huaxi School of Public Health, University of Sichuan, Chengdu, China. E-mail: zzmao666@yahoo.com
    Search for more papers by this author

    1. School of Public Health, University of California, Berkeley, USA
    Search for more papers by this author

Zhengzhong Mao, Department of Health Economics, Huaxi School of Public Health, University of Sichuan, Chengdu, China. E-mail: zzmao666@yahoo.com


Background. Secondhand smoke (SHS) exposure harms pregnant women and the fetus. China has the world's largest number of smokers and a high male smoking prevalence rate. Objective. To compare exposure to SHS among rural and urban Chinese non-smoking pregnant women with smoking husbands, and analyze factors associated with the level of SHS exposure and hair nicotine concentration Setting. Sichuan province, China. Population. In all 1,181 non-smoking pregnant women with smoking husbands recruited from eight district/county Women and Children's hospitals. Methods. The women completed a questionnaire in April and May 2008. Based on systematic sampling, 186 pregnant women were selected for sampling the nicotine concentration in their hair. Ordinal logistic regression analysis was conducted to examine correlates with self-reported SHS exposure (total and at home); linear regression was conducted for the sub-sample of hair nicotine concentrations. Main outcome measures. Secondhand smoking exposure rates, hair nicotine levels. Results. About 75.1% of the non-smoking pregnant women with smoking husbands reported regular SHS exposure. The major source of exposure was through their husband. In the multivariate analysis, the risk of greater SHS exposure (total and at home) and hair nicotine concentration was increased for women who were rural, had a husband with greater cigarette consumption, less knowledge about SHS, less negative attitudes about SHS, and no smoke-free home rules. Conclusions. The high prevalence rate of SHS exposure suggests that it is important for non-smoking pregnant women, especially rural women, to establish smoke-free home rules and increase knowledge and negative attitudes towards SHS.


Secondhand smoke (SHS), for which there is no risk-free level of exposure, causes disease in non-smokers (1). Non-smokers are at increased risk of developing cancer, cardiovascular disease and chronic respiratory illness when exposed to SHS (1). For non-smoking pregnant women, SHS exposure can also cause harmful effects to the fetus: low birthweight, fetal death, preterm delivery and spontaneous abortion (2–4).

China's one-child family policy (5) and cultural emphasis on family present a unique public health challenge to examine how the knowledge, attitudes, and exposure to SHS are associated among non-smoking pregnant women. China has the world's largest number of smokers (350 million with 66% male, 3% female) and half of the non-smoking population (540 million) report regular exposure to SHS (6). National estimates report 54.6% of all women as exposed to SHS, with 90% in the home, and 60% of non-smoking pregnant women having husbands who are regular smokers (6, 7). The harmful effects of SHS exposure have been well-documented in studies conducted in China on the fetus (8–13) (such as pregnancy loss, low birthweight, preterm delivery, fetal death) and children's long-term risk for severe infectious morbidity (14).

Research in two Chinese cities has previously reported on pregnant women's exposure to SHS. In Guangzhou, a study of 1,449 never-smoking pregnant women who made their first prenatal visit to one hospital during 1996–1997 showed that 60.2% had a husband who smoked regularly during the woman's gestation. These women were more exposed to SHS than those with non-smoking husbands at home (71 vs. 33%), in public places (77 vs. 66%) and at work (60 vs. 50%) (15). In Shanghai, 950 new mothers with infants were recruited from three communities in 2005–2006 and the mothers interviewed about SHS exposure before, during and after pregnancy (16). The prevalence of SHS exposure was 55.9% before pregnancy, which decreased significantly to 41.9% during the pregnancy, but the amount and time of exposure changed little for those who continued to be exposed (16). Unfortunately, previous research in China is limited in that these studies were only confined to people in two major cities, which means there is still an inadequate understanding of SHS exposure among rural pregnant women. SHS exposure in rural areas is much higher than that in cities (54.0 vs. 49.7%) (7) and the rural population make up more than half of China's entire population. In addition, previous research did not identify correlates of SHS exposure, which could be of use for the development of interventions. Finally, previous research only measured exposure by survey recall, which may be more likely to be subject to misclassification of exposure (17).

The objectives for this study of rural and urban non-smoking pregnant women with smoking husbands in China's Sichuan province were to compare SHS exposure and identify associated factors. Two research methods were used in parallel. The first involved a questionnaire assessment and the second was a hair nicotine measurement taken from a systematically sampled subset of participants to estimate their long-term exposure to SHS (18). It was hypothesized that the pregnant women with smoking husbands in rural areas would have a higher SHS exposure rate, higher hair nicotine levels, less knowledge and negative attitudes about SHS, and less smoke-free home rules than respondents in urban areas.

Material and methods

Study population

This cross-sectional study was conducted in April and May 2008. Fu's study showed the prevalence of SHS exposure during pregnancy to be 41.9%, which was used to determine the required sample size (16). For a 5% level of significance, 555 women were needed. To ensure a representative sample size, eight district and county women and children's hospitals in Sichuan province (Chenghua District, Jintang County, Longchang County, Dongpo District, Renshou County, Da'an District, Longquanyi District, Jingyan County) were selected by a computerized random number generator for 32 district/county Women and Children's hospitals, which all had an annual number of live births over 1,000. All pregnant women who visited the hospitals to set up their prenatal files during the research period and who met the inclusion criteria, were invited to participate. Inclusion criteria were: (1) a pregnant woman who had never smoked before, (2) the pregnant woman's husband was a smoker and they lived together during her pregnancy, and (3) she was willing to participate. The study was approved by the medical science ethical committee of Sichuan University, and all subjects provided informed consent before participation. Each participant received baby products valued at 10–20 Chinese Yuan for participating.


Based on a previous study by Lee (19) about non-smoking pregnant women in China exposed to SHS, the questionnaire consisted of several sections: demographics, knowledge and attitudes towards SHS, household smoking rules and self-reported exposure at different locations. Demographic questions included age, education level, employment status, family net income/month, rural vs. urban region (classified according to the Hukou system, a household registration system used in China) and husband's daily cigarette consumption. SHS exposure was defined as exposure to another person's tobacco smoke for at least 15 minutes daily for more than one day every week (20), and participants were asked about source (home, work, public places) and daily average duration of exposure during the last week. Knowledge about SHS was measured by questions for three domains allowing for multiple responses, based on Lee's study (19): definition and components of SHS, diseases caused by SHS, and harmful effects on pregnancy and the fetus caused by SHS. The questions about the definition and components of SHS were (1) ‘Is SHS the inhalation of smoke from tobacco products used by others?’ (2) ‘What kind of harmful ingredients does SHS contain?’ The question about diseases caused by SHS was ‘What kinds of disease are related to SHS?’ The questions about maternal and fetal diseases caused by SHS were ‘What kinds of effect would SHS cause to yourself? What kinds of effect would SHS cause to the fetus?’ The answer to every possible response was either ‘yes’ or ‘no’, and was scored +1 for a correct answer and 0 for an incorrect or missing answer. The score for each domain mentioned above were 7, 5 and 12, respectively (total score: 24). Participant attitudes when exposed to SHS during their pregnancy were categorized as “strongly dislike”, ‘dislike’ and “indifferent” (19), finding that women reported confusion with the more standard 5-point Likert scale including ‘like’ and ‘strongly like’ for this topic. Household smoking rules were divided into three categories: (1) smoking was allowed in all parts of the home, (2) smoking was allowed in some parts of the home, and (3) smoking was not allowed in any part of the home. The questionnaire was pilot-tested among 50 non-smoking pregnant women for validity, reliability, linguistic appropriateness, and cultural differences in the field before being finalized. The questionnaire was distributed and collected by a trained doctor or nurse, but self-administered by the participant.

Hair nicotine measurements

Hair nicotine, a long-term biomarker of smoke exposure, was measured in a systematically sampled subset of participants. The random starting point was 5 with additional participants selected who were multiples of ‘five’. Hair strands were cut at the scalp from the back of the head around the occipital bone. The sample was then trimmed to 1.0–1.5 cm from the root end, totaling approximately 1.0 g in weight. Because human hair grows almost 1.0 cm per month, the 1.0–1.5 cm reflects SHS exposure in the past month (21). Hair samples were immediately placed in a labeled envelope that was sealed in a plastic bag for storage and transportation. This analytical method was an adaptation of the procedure developed by Kintz (22). The hair sample was digested by 1.5 mol/l sodium hydroxide solution. The nicotine in the hair sample was extracted with a mixed solvent of dichloromethane-methanol (3:1). An aliquot of the extraction solution was vaporized with nitrogen flow and then methanol was added to dissolve the sample. For quantification, a nicotine standard (Sigma Chemical Company, St. Louis, MO, USA) was used. The analytes were tested with capillary gas chromatography at a laboratory in Sichuan University.

Statistical analysis

Statistical analysis was performed with SPSS 13.0 software (Chicago, IL, USA). Results are reported as median, geometric mean or numbers with percentages. Categorical data were analyzed using chi-squared analyses, and continuous variables using independent Student's t-tests. Spearman's rank correlation was used to assess the relation between self-reported exposure and nicotine levels. For the ordinal logistic regression analyses, the total self-reported exposure and the exposure at home per day was divided into no exposure, 15–59 minutes exposure and over 60 minutes exposure. The analysis was set up to examine associations with higher amounts of exposure. The independent variables included the women's age, educational background, working situation, monthly family net income, urban vs. rural status, husband's daily cigarette consumption, knowledge of SHS, attitudes towards SHS and household smoking rules. Next, multivariate linear regression model was used to examine factors associated with log hair nicotine, among the subset of women sampled. The independent variables were the same as with the ordinal logistic regression. There was no strong relation between independent variables in the regression models. All analyses were two-tailed with a significance level of p < 0.05.


A total number of 1,263 eligible pregnant women visited the hospitals to set up their prenatal files, and 1,181 of them (93.5%) took part and completed questionnaires which were utilized in the analysis. For hair nicotine measurements, 236 pregnant women were systematically approached from among all participants, but only 186 agreed to have their hair sampled. Neither there was any difference in demographic characteristics between the refusers for the study and participants, nor refusers for the hair sample and other women in the participant sub-sample (data not shown). Of the 1,181 non-smoking pregnant women, 784 were rural (66.4%) and 397 were urban (33.6%). All of the women were married and lived with their husband during their pregnancy.

Demographic characteristics are summarized in Table 1. The mean age of the women was 25.6 (SD = 4.5) years. Approximately half had a middle school level of education or less, were employed and had a family net income per month of 1,000–3,000 Yuan. Compared to the urban women, the rural women were younger, had a lower employment rate, a lower education level, and less family net income (p < 0.001). Among currently smoking husbands, over half smoked >10 cigarettes daily, and there was no significant difference between rural and urban regions (p = 0.26).

Table 1. Demographics and knowledge and attitudes about SHS of non-smoking pregnant women living with a smoking husband in Sichuan province, China.
CharacteristicAll (n = 1,181)Rural women (n = 784)Urban women (n = 397)p
Age (years) (mean ± SD)25.6 ± 4.525.0 ± 4.526.8 ± 4.3<0.001
 Middle school or less than middle school63353.652867.310526.4 
 High school38132.321427.316742.1 
 College or higher16714.1425.412531.5 
Working situation      <0.001
Family net income per month (RMB)      <0.001
 Under 1,00032927.927034.45914.9 
 Over 3,00030025.414017.916040.3 
Husband's daily cigarette consumption      0.26
 Under 524821.015920.38922.4 
 Over 2012010.28611.0348.6 
Household smoking ban      <0.001
 Smoking was allowed in all parts of the home64254.446559.317744.6 
 Smoking was allowed in some parts of the home37131.421727.715438.8 
 Smoking was not allowed in any part of the home16814.210213.06616.6 
Scores of SHS knowledge (mean ± SD)5.5 ± 3.94.9 ± 3.76.6 ± 3.8<0.001
Attitudes towards SHS      0.46
 Strongly dislike60251.039350.120952.6 

The mean summary score for knowledge about SHS was very low. Compared to urban women, the rural pregnant women scored lower as analyzed by the Student's t-tests. For attitudes towards SHS exposure during pregnancy, the percentage of non-smoking pregnant women who chose ‘strongly dislike’ was 51.0%, ‘dislike’ was 42.4%, and ‘indifferent’ was only 6.6%. No significant difference was found between the attitudes of rural and urban women (p = 0.46) (Table 1).

The percentage of non-smoking pregnant women who reported smoke-free home rules was only 14.2%, however, 31.4% of the women reported smoking was allowed in some parts of the home, while 54.4% of the women reported smoking was allowed in all parts of the home. Rural and urban women differed in this regard (p < 0.001).

For SHS exposure (Table 2), 75.1% of non-smoking pregnant women reported being exposed to SHS for at least 15 minutes daily for more than one day a week, with almost half of these women reporting exposure for 15–59 minutes per day and the other half reporting exposure for one hour or more per day during the past week. Of the women reporting regular exposure, the percentage of SHS exposure at home was 57.7%. The main source of exposure was from husbands (more than half of the participants), followed by public places (21.4%), family members other than the husband (11.9%), colleagues at work (11.6%), and other persons (12.7%). Rural pregnant women reported more SHS exposure than did the urban women (p < 0.001), especially with regard to exposure at home and from the husband (p < 0.001).

Table 2. Self-reported SHS exposure at different locations for non-smoking pregnant women living with a smoking husband in Sichuan province, China.
ExposureAll (n = 1,181)Rural women (n = 784)Urban women (n = 397)p
Past week total average duration of the exposure (per day)      <0.001
 No exposure29424.917322.112130.5 
 15–59 minutes46639.531640.315037.8 
 ≥ 60 minutes42135.629537.612631.7 
Past week total average duration of the exposure at home (per day)      <0.001
 No exposure49942.327535.122456.4 
 15–59 minutes45638.633342.512331.0 
 ≥ 60 minutes22619.117622.45012.6 
Past week average duration of exposure from husband (per day)      <0.001
 No exposure55046.633843.121253.4 
 15–59 minutes48441.033442.615037.8 
 ≥ 60 minutes14712.411214.3358.8 
Past week average duration of exposure from other family members (excluded husband) (per day)      0.065
 No exposure104146.668086.736190.9 
 15–59 minutes7741.0607.7174.3 
 ≥ 60 minutes6312.4445.6194.8 
Past week average duration of exposure from colleagues (per day)      0.54
 No exposure104488.469488.535088.2 
 15–59 minutes675.7415.2266.5 
 ≥ 60 minutes705.9496.3215.3 
Past week average duration of exposure from other people at work places (per day)      0.17
 No exposure103887.968387.135589.4 
 15–59 minutes746.3486.1266.5 
 ≥ 60 minutes695.8536.8164.0 
Past week average duration of exposure from public place (per day)      0.46
 No exposure92878.660877.632080.6 
 15–59 minutes17414.712215.65213.1 
 ≥ 60 minutes796.7546.9256.3 

In the ordinal logistic regression model (Table 3), a non-smoking pregnant woman was more likely to report higher total SHS exposure if she lived in a rural area, had a husband with higher cigarette consumption, did not have smoke-free home rules, had less knowledge about SHS, and a less negative attitude against SHS exposure. Similar associations were found with the model for higher home SHS exposure.

Table 3. Ordinal logistic regression model of factors associated with self-reported total or home secondhand smoke exposure for non-smoking pregnant women living with a smoking husband in Sichuan province, China (n = 1,181).
Total secondhand smoke exposureSecondhand exposure at home
OR95% CIpOR95% CIp
Age (years)0.980.96,, 1.020.75
 Rural1.961.34, 2.84<0.0012.271.50, 3.44<0.001
 Middle school or less than middle school0.780.43, 1.440.430.620.33, 1.170.14
 High school1.110.68, 1.820.670.880.52, 1.500.64
 College or higher    
Working situation      
 Unemployed0.840.67,, 1.210.19
Husband's daily cigarette consumption      
 Under 50.100.06, 0.16<0.0010.040.02, 0.06<0.001
 5–90.160.10, 0.26<0.0010.100.06, 0.16<0.001
 10–140.250.15, 0.40<0.0010.210.14, 0.34<0.001
 15–190.230.38, 0.14<0.0010.230.14, 0.36<0.001
 Over 20    
Smoking home rules      
 Smoking was allowed in all parts of the home3.011.95, 4.64<0.0013.912.41, 6.33<0.001
 Smoking was allowed in some parts of the home2.011.36, 2.97<0.0011.571.02, 2.430.04
 Smoking was not allowed in any part of the home    
Family net income per month (RMB)      
 Under 9990.920.66, 1.280.610.920.65, 1.310.27
 1,000–2,9990.960.72, 1.280.790.990.74, 1.350.30
 Over 3,000    
Scores of SHS knowledge0.920.90, 0.95<0.0010.870.84, 0.90<0.001
 Indifferent2.181.32, 3.610.0022.421.47, 3.970.001
 Dislike1.170.93, 1.480.191.150.90, 1.480.27
 Strongly dislike    

Concentration of nicotine in the hair samples ranged from 0.52 to 25.57 ug/g. Self-reported (minutes/day) SHS exposure was correlated with hair nicotine concentration (Spearman's r = 0.3; p < 0.0001). The geometric mean of nicotine concentration in the hair of rural pregnant women (3.39 ug/g) was significantly higher than that of urban women (2.63 ug/g) (p = 0.002). In the multivariate linear regression models (Table 4), factors negatively associated with increasing log concentration of hair nicotine levels included having more knowledge about SHS; factors positively associated were living in a rural area, having less negative attitudes towards SHS exposure, allowing smoking in all parts of the house, and a husband with higher cigarette consumption. The final model including all these variables accounted for 45.0% of the variance in concentration of nicotine in the hair.

Table 4. Multivariate linear regression model fitted on log mean of nicotine concentration in hair for non-smoking pregnant women living with a smoking husband in Sichuan province, China (n = 186).
Predictor variableB coefficient95% CIt Statisticp
  1. Note: R2 = 0.45, F = 8.00, p = <0.001.

Age (years)−0.004−0.01, 0.00−0.830.41
 Rural0.160.07, 0.253.450.001
 Middle school or less than middle school0.002−0.23,
 High school−0.04−0.18, 0.10−0.500.62
 College or higher   
Working situation    
 Unemployed0.05−0.03, 0.131.320.19
Husband's daily cigarette consumption    
 Under 5-0.28−0.45, -0.11−3.260.001
 5-9-0.19−0.35, -0.03−2.330.02
 10-14-0.11−0.27, 0.06−1.310.19
 15-19-0.09−0.25, 0.07−1.070.29
 Over 20   
Smoking home rules    
 Smoking was allowed in all parts of the home0.190.06-0.312.970.003
 Smoking was allowed in some parts of the home0.46−0.09, 0.180.690.49
 Smoking was not allowed in any part of the home   
Family net income per month (Yuan)    
 Under 9990.22−0.06, 0.501.530.13
 1,000-2,9990.09−0.10, 0.280.970.33
 Over 3,000   
Scores of SHS knowledge-0.02−0.03,−0.01−3.250.001
 Indifferent0.210.06, 0.362.710.007
 Dislike0.100.02, 0.212.340.021
 Strongly dislike   
 Constant0.26-0.23, 0.751.050.30


This study is the first to report the exposure to SHS in non-smoking pregnant women in China on a provincial level, and is novel for including a large rural population and hair nicotine analysis. The high prevalence of SHS exposure for the non-smoking pregnant women in our provincial study (75%) is consistent with previous literature described 10 years ago in the city of Guangzhou (15). Although one study in the USA reported that 16.4% of non-smoking singleton pregnancies had SHS exposure during pregnancy (23), certain subpopulations even in developed countries like the USA may have high SHS exposure rates. Among pregnant women in New Haven, CN, USA, 52% of non-smokers had been classified as having had recent SHS exposure according to their urinary cotinine levels (24). Another study examined correlates of SHS avoidance in a population of African-American pregnant non-smokers who lived with smokers and reported 73% of the women's salivary cotinine levels exceeded the passive smoking cut-off of 10 ng/ml (25).

Rural participants reported higher SHS exposure than urban participants, and this was confirmed by the hair nicotine analysis, even after controlling for demographic factors, knowledge of, and attitudes towards, SHS. The persistent association between rural non-smoking pregnant women and higher SHS exposure, compared to urban women, may have resulted from demographic differences like education, but also may reflect sociocultural differences. Yang's study showed that rural residents in China's western region had low levels of knowledge about the health harms of SHS exposure (26). In addition, Ma's research found multiple barriers that sustain smoking in China from 80 focus groups and 30 in-depth interviews in three rural counties in Jiangxi, Henan and Sichuan provinces; for example, tobacco was important in social and cultural interactions (27). In the face of these social norm differences, rural non-smoking pregnant women may suffer more exposure to SHS than urban women.

Our study found that a husband with higher cigarette consumption and a lack of smoke-free home rules were two factors significantly associated with higher levels of self-reported SHS exposure (total and home) and hair nicotine concentrations. These findings are consistent with other literature. Among pregnant African-American non-smokers, the odds of SHS avoidance were increased among women who reported household smoking bans and that no/few family members/friends smoked (28). Wong's study showed that among children, the information about level of household smoking restriction, parental smoking status, and number of cigarettes smoked per day accounted for approximately 45% of the variance in cotinine concentration (29).

Establishing and enforcing smoke-free home rules may be a strategy for protecting non-smoking pregnant women and their fetuses and promoting the husband's smoking cessation. In Taiwan, overall avoidance of SHS by non-smoking pregnant women was associated with higher self-efficacy, higher educational levels, and smoke-free home rules (30). Blackburn et al. found that banning smoking at home was associated with significant reductions in urinary cotinine to creatinine ratio in infants (31). In California, smokefree home rules are associated with Asian-American men, particularly recent immigrants, being former smokers (32). Empowerment of women to enforce policies is important since, in California, lower-educated compared to higher-educated Asian-American women (Chinese and Korean) have similar rates of smoke-free home rules and indoor work policies but report greater recent exposure, despite controlling for knowledge about SHS harms (33).

Our study suggests that stronger knowledge and more negative attitudes about SHS are important factors associated with less self-reported SHS exposure, all of which are associated with lower long-term exposure levels (as measured by our sub-sample of participants with hair nicotine). Thus, non-smoking pregnant women may play an important role in protecting their own health. To date, two intervention studies have examined empowering non-smoking pregnant women to protect the health of themselves and their fetus. Loke and Lam conducted a randomized controlled trial in the city of Guangzhou demonstrating that simple advice from the obstetrician to the non-smoking pregnant women helped more smoking husbands quit in one month (34). In 2007, Lee conducted a pilot intervention study which demonstrated that educating non-smoking pregnant women about SHS and positive communication skills led to increased self-reports of assertive actions against SHS exposure (19). This intervention is currently being tested in an ongoing larger trial with the participants from this study.

The second most frequent reported source of exposure to SHS was from being in public. However, participants reported much less assertive actions against SHS exposure in public compared to their husbands. This may reflect an environmental difference in that people have less control over their public environment than their home environment, but also a cultural difference in that individual attitudes matter less than social norms. Smoke-free public regulations have been effective in reducing exposure (1). At present, China is implementing the World Health Organization's Framework Convention on Tobacco Control, which promotes smoke-free regulations. While the 2008 Olympics led to widespread smoke-free policies in Beijing, the enforcement and dissemination of such policies need to be continued.

Overall, the non-smoking pregnant women in this study have very little knowledge about SHS. This may reflect that half of the women in our study did not have high levels of education (middle school or less). In Lee's research, 88.3% of non-smoking pregnant women had an education level of high school or higher and had a much better understanding about SHS than women in our research (19). Future media and educational campaigns in China should consider increasing knowledge about SHS health harms and promoting smoke-free social norms based on negative reactions. Such campaigns are rarely carried out in China currently but have had success in California, where in-language smoke-free media campaigns have led to non-smokers speaking Asian languages seeking help for smokers (35). With China's ‘one-child’ policy, pregnant women would benefit if they knew more about the hazards of SHS, such as preterm labor, abortion, fetal death, and low birthweight. The channels that could be used to spread knowledge include special educational programs on health, consultation provided by obstetricians and obstetric nurses, and mass media, such as internet and newspaper.

There are three main limitations to this study. First, the participants were from Sichuan province and may not reflect the whole of China. The female smoking prevalence in Sichuan is 4% (36), which is a little higher than the national prevalence of 3%, and attitudes and behavior may differ. Second, the study only sampled women who sought prenatal health care in Women's and Children's hospitals and excluded those who visited general hospitals. Thus, generalization of the results may be somewhat limited. Third, due to limited resources, hair nicotine was not measured of all participants, although a systematic sampling protocol was in place.

In summary, this study shows that non-smoking pregnant women in China who lived with a smoking husband were highly exposed to SHS, especially from their husbands. In addition, non-smoking pregnant women have inadequate knowledge on the harms of SHS. These findings are particularly significant for rural women. In the multivariate analysis, the risk of greater SHS exposure (total and at home) and hair nicotine concentration was increased for women who were rural, had a husband with greater cigarette consumption, less knowledge about SHS, less negative attitudes about SHS, and no smoke-free home rules. Results highlight the importance of encouraging smoke-free home rules and taking measures to empower the non-smoking pregnant women in reducing the exposure of themselves and their fetuses.


This study was funded by NIH Fogarty International Center grant TW05938. The authors are grateful to the medical personnel who collected the samples. The authors would like to thank Dr. Anita H. Lee, Center for International Tobacco Control, Public Health Institute, Berkeley and Ms. Tingting Yao, doctoral student of Sichuan University, for their related pilot research.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.