Abstract

Results of studies on paternal smoking and spontaneous abortions have been inconsistent. The authors examined the effect of paternal smoking on the risk of pregnancy loss in a prospective cohort of 526 newly married, nonsmoking, female textile workers in China between 1996 and 1998. Upon stopping contraception, subjects provided daily urine specimens and records of vaginal bleeding for up to 1 year or until clinical pregnancy. Daily urinary human chorionic gonadotropin was assayed to detect conception and early pregnancy losses, and pregnancies were followed to detect clinical spontaneous abortions. Subjects were grouped by the number of cigarettes that husbands reported smoking daily: nonsmokers (group 1, n = 216), fewer than 20 cigarettes (group 2, n = 239), and 20 or more cigarettes (group 3, n = 71). Compared with that for group 1, the adjusted odds ratio of early pregnancy loss of any conception for group 2 was 1.04 (95% confidence interval (CI): 0.67, 1.63) and for group 3 was 1.81 (95% CI: 1.00, 3.29). The adjusted hazard ratio of conception for group 2 was 0.90 (95% CI: 0.70, 1.18) and for group 3 was 0.96 (95% CI: 0.66, 1.39), while the adjusted hazard ratio of clinical pregnancy for group 2 was 0.93 (95% CI: 0.72, 1.20) and for group 3 was 0.78 (95% CI: 0.55, 1.12). The authors conclude that heavy paternal smoking increased the risk of early pregnancy loss through maternal and/or paternal exposure.

Received for publication June 17, 2003; accepted for publication December 4, 2003.

In many regions of the world, the prevalence of tobacco use is much higher among men than women. In China, for example, a 1996 national study of smoking prevalence found that 70 percent of men but only 4 percent of women currently smoked. However, 70 percent of nonsmoking Chinese women between the ages of 20 and 50 years reported exposure to passive smoke. Among exposed women, 72 percent reported that their passive smoke exposure occurred daily (1). Furthermore, studies based upon both self-reports and biologic indicators of exposure have shown consistently that the exposure of pregnant women to passive smoke is widespread (2) and that the smoking status of a woman’s partner is the strongest predictor of her exposure to passive smoke (3). Among women reporting passive smoke exposure in the national Chinese survey, 82 percent reported exposure at home compared with 28 percent in public places and 19 percent at work (1). It is important to elucidate the effects of paternal smoking on pregnancy because, even if the magnitudes of effects are modest, the adverse impacts on the public’s health will be widespread with exposure so prevalent.

Although previous studies have shown that active maternal smoking increased the risk of spontaneous abortion (46), results of investigations of increased risk due to maternal exposure to passive smoke have been inconsistent. Most have shown no evidence of an effect on the risk of spontaneous abortion (4, 711). Some authors have suggested that inconsistent findings in previous studies of spontaneous abortion and passive smoke might have resulted from a failure to assess early pregnancy losses occurring before clinical detection of pregnancies (12). We are unaware of any previous studies that have directly investigated the effect of paternal smoking on early pregnancy loss. However, to define accurately all conceptions and pregnancy losses in a population, researchers must follow a cohort of women prospectively from the beginning of cycles without contraception until occurrence of the endpoints of interest. The detection of early pregnancy loss, which is not clinically apparent, requires a sensitive and specific test for pregnancy such as the assay of urinary human chorionic gonadotropin (hCG). Without the use of an appropriate biomarker, only clinical spontaneous abortion would be observable because early pregnancy loss would not be detectable.

The development of sensitive and specific urinary hCG assays has made it possible to detect pregnancy close to the time of implantation (13). With the hCG assay, several studies have shown that approximately one third of all conceptions detected by the hCG assay failed to survive to delivery and that over two thirds of pregnancy losses occurred before clinical detection of pregnancy (14, 15). The use of urinary hCG as a biomarker provides an opportunity to investigate the effect of paternal smoking on early pregnancy loss.

We investigated the effect of paternal smoking on pregnancy loss using a large prospective cohort of women who participated in a reproductive health study in Anhui, China. We report the effect of husbands’ current smoking on time to conception and clinical pregnancy and on pregnancy losses occurring before and after clinical detection of pregnancy.

MATERIALS AND METHODS

Study population

This is part of a large prospective study of reproductive health conducted from 1996 to 1998 in textile mills in Anhui, China. The Institutional Review Board of the Harvard School of Public Health and collaborating Chinese institutions approved our study protocols. We obtained written informed consent from each woman and her husband.

We previously reported a detailed description of the study population and data collection methods (15). Briefly, the eligibility criteria for women in the field enrollment were as follows: 1) full-time employment, 2) aged 20–34 years, 3) newly married, and 4) had obtained permission to have a child. All the women were nulliparous. Women were excluded if they 1) were already pregnant before enrollment, 2) had tried unsuccessfully to get pregnant for 1 year or longer at any time in the past, or 3) planned to quit/change jobs or move out of the city over the 1-year course of follow-up.

Of 1,006 newly married women who were screened (more than 90 percent of newly married women employed at the mill), 961 met eligibility requirements and agreed to enroll. We excluded 435 enrolled women from this analysis because they did not collect daily urine (n = 121), did not begin collecting urine soon after stopping contraception (n = 53), never stopped using contraception (n = 95), became pregnant because of contraceptive failure (n = 78), were lost to follow-up (n = 8), withdrew shortly after enrollment (n = 27), had inadequate diary data (n = 12), did not have baseline data about husband’s smoking (n = 34), or had menstrual cycles less than 21 or greater than 40 days in length (n = 7). The characteristics of the excluded women were similar to those who were included (15). This report includes 526 women who intended to conceive, who began recording diary entries and collecting daily urine samples immediately after stopping contraceptive use, and who had adequate diary and hCG data. All of the women included in our analysis had never been smokers.

After obtaining informed consent, we interviewed eligible women and their husbands separately using questionnaires collecting baseline information on menstruation, contraceptive use, reproductive history, sociodemographic characteristics, smoking (including number of cigarettes currently smoked each day), alcohol use, and environmental and occupational exposures. We defined smokers as those who had smoked at least one cigarette/day continuously for one-half year or who had smoked at least 10 packs of cigarettes in total (regardless of how long the time period was during which they smoked). Former smokers were not current smokers and had been a smoker previously. We measured weight and height using standard methods. Beginning from the date of stopping use of contraceptive methods, each woman kept a daily diary to record contraceptive use, sexual intercourse, vaginal bleeding, medication, and medical conditions and collected a daily first-morning urine specimen for hCG assay. We collected daily diary information and first morning urine specimens for up to 12 months or until a pregnancy was clinically confirmed. We monitored women during ensuing pregnancies (or up to 1 year after beginning to attempt pregnancy if there was no pregnancy) and recorded all pregnancy outcomes.

Laboratory assay of urinary human chorionic gonadotropin

We analyzed urine specimens for hCG with the immunoradiometric assay developed by O’Conner et al. (16) using a combination of capture antibodies for hCG-free beta subunit and hCG beta core fragment (B204) and for the intact hCG molecule (B109). This assay was highly sensitive and specific. The lowest hCG concentration detectable by the assay was 0.01 ng per ml (1 mIU = 0.2 ng) (17). The cross-reaction of the assay with either intact luteinizing hormone or luteinizing hormone-free subunit was less than 1 percent. We analyzed and tested all urine specimens from each woman during a single run of the assay. We assayed each urine specimen in duplicate during the window of –10 to +5 days of a menstrual cycle and repeated the assays if the discrepancy between duplicates was greater than threefold. For pairs of duplicate assays with less than a threefold difference, we used the geometric mean of the assays in our analyses. We used the method of Jaffe as described by Husdan and Rapoport (18) to measure urine creatinine levels, and we normalized hCG values to creatinine to adjust for urine concentration. As reference values, we determined the levels of hCG from 67 nonconception cycles of 37 control women who were married but using contraception (n = 4), not married (n = 23), or married but not cohabitating with their husband (n = 10) (15).

Major outcomes and methods of evaluation

The major outcomes and methods of evaluation are described as follows: 1) conception: a conception detected by urinary hCG assay (see below); 2) clinical pregnancy: any pregnancy that lasted 6 weeks or more (≥42 days) after the onset of the last menstrual period and that was confirmed by hCG assay; 3) early pregnancy loss: pregnancy loss (detected by urinary hCG assay) occurring less than 6 weeks (<42 days) after the onset of the last menstrual period; 4) clinical spontaneous abortion: pregnancy loss occurring 6 weeks or more (≥42 days) but no later than 20 weeks after the onset of the last menstrual period; and 5) total pregnancy loss (early pregnancy loss plus clinical spontaneous abortion): pregnancy loss occurring no later than 20 weeks after the onset of the last menstrual period.

Statistical methods

One of the key outcomes of this investigation was early pregnancy loss, which we determined by observing daily urinary hCG values. To distinguish normal variation from a true hCG rise due to conception and to address missing hCG values, we used Bayesian methods (19, 20) to model daily conception status among all the female subjects, including control women who did not conceive. In a previous report (15), we showed that this model was 100 percent sensitive and specific for those cycles in which the conception status was observable; that is, the probability of conception was 0.0 in all control cycles and 1.0 in all cycles with conception leading to clinical pregnancy.

To investigate the exposure-response relation between paternal smoking and pregnancy losses, we modeled the amount of husband’s smoking as an ordinal variable defined by the number of cigarettes that he reported smoking each day at the time of the baseline questionnaire. Reported numbers of cigarettes smoked by the husbands showed strong digit preferences for multiples of five, especially 10 and 20. Cigarettes in China are generally sold in packs of 20. Because of digit preference and to balance the sample size in each category, we first conducted our analyses using four ordinal categories of smoking: none, low (fewer than 10 cigarettes/day), medium (10–19 cigarettes/day), and high (≥20 cigarettes/day). Because the effects of husbands’ smoking were found only among those in the highest smoking group, we combined the low and medium smoking groups and repeated the analyses. We present the results here using three ordinal categories of smoking: none, fewer than 20 cigarettes/day, and 20 or more cigarettes/day.

Using women whose husbands did not smoke as a reference group, we estimated the relative odds of early pregnancy loss and total pregnancy loss (early pregnancy loss or clinical spontaneous abortion) among women whose husbands smoked fewer than 20 cigarettes/day and 20 or more cigarettes/day. We initially calculated the relative odds of early pregnancy loss and total pregnancy loss in the first conception cycle and then repeated both analyses three times using the first two, first three, and all conceptions. In models that included multiple conceptions by some women, we estimated standard errors using generalized estimating equations to accommodate correlations in pregnancy losses among conceptions (21). We used Cox proportional hazards methods (22) to estimate the relative hazards of conception and clinical pregnancy for each smoking group using women whose husbands did not smoke as a reference group.

We adjusted for important covariates in this analysis including both the wife’s and the husband’s ages (tertiles), education (binary, high school or above), perceived life stress (binary, response of moderate or high to the question, “How would you describe the level of stress in your daily life?”), and exposures to dust (binary) and noise (binary); the husband’s use of alcohol (binary), previous smoking (binary), and exposure to any toxins including solvents, metals, gases, fumes, or acids (binary); and the wife’s body mass index (binary, body mass index of <19 or ≥19) and tea drinking (binary).

RESULTS

About 45 percent of the husbands included in this report smoked fewer than 20 cigarettes per day, and 13 percent smoked 20 or more cigarettes per day at the time of the baseline survey. As shown in table 1, women in the three husbands’ smoking groups were similar in age, height, weight, body mass index, and toxin exposures. Compared with the other two groups, women whose husbands smoked fewer than 20 cigarettes per day had lower education, and women whose husbands smoked 20 or more cigarettes per day had higher dust and noise exposures and less moderate-to-high stress and were more likely to drink tea. None of the women drank alcohol. Husbands in the three smoking groups were similar in height, weight, and body mass index. Compared with the other groups, husbands who smoked fewer than 20 cigarettes per day had higher education and higher dust, noise, and toxin exposures. Husbands who smoked 20 or more cigarettes per day were older, less educated, and more likely to drink tea and alcohol and had lower perceived stress.

Table 2 shows that similar percentages of women in each smoking group achieved at least one conception (nonsmoking: 95 percent; <20 cigarettes/day: 95 percent; and ≥20 cigarettes/day: 96 percent) and clinical pregnancy (nonsmoking: 90 percent; <20 cigarettes/day: 92 percent; and ≥20 cigarettes/day: 89 percent). There were slightly fewer women in the ≥20 cigarettes/day group compared with the other two groups who had clinical spontaneous abortion (nonsmoking: 10 percent; <20 cigarettes/day: 10 percent; and ≥20 cigarettes/day: 6 percent). Among those who achieved clinical pregnancy without clinical spontaneous abortion, there were slightly fewer women in the ≥20 cigarettes/day group compared with the other two groups who experienced no early pregnancy loss (nonsmoking: 84 percent; <20 cigarettes/day: 81 percent; and ≥20 cigarettes/day: 76 percent) and slightly more who experienced three or more early pregnancy losses (nonsmoking: 0 percent; <20 cigarettes/day: 2 percent; and ≥20 cigarettes/day: 5 percent).

Table 3 shows that, among the 94 percent of subjects who conceived at least once, the group of women whose husbands smoked ≥20 cigarettes/day had the highest prevalence of early pregnancy loss in the first conception (nonsmoking: 22 percent; <20 cigarettes/day: 20 percent; and ≥20 cigarettes/day: 29 percent), the lowest prevalence of clinical spontaneous abortion (nonsmoking: 8 percent; <20 cigarettes/day: 10 percent; and ≥20 cigarettes/day: 4 percent), and the highest prevalence of total pregnancy loss (nonsmoking: 29 percent; <20 cigarettes/day: 30 percent; and ≥20 cigarettes/day: 34 percent). Second conceptions had the same pattern of prevalence of early pregnancy loss, clinical spontaneous abortion, and total pregnancy losses in the three smoking groups as did the first conceptions. Among the relatively few third conceptions that occurred (n = 28), the group of women whose husbands smoked ≥20 cigarettes/day had the highest prevalence of early pregnancy loss and total pregnancy loss.

Table 4 shows the relative odds of early pregnancy loss by husband’s smoking status in the first conception, the first two conceptions, the first three conceptions, and all conceptions. Compared with those for women whose husbands did not smoke, the adjusted odds of early pregnancy loss in the first conception cycle for women whose husbands smoked <20 cigarettes/day were 0.81 (95 percent confidence interval (CI): 0.49, 1.33) and, for women whose husbands smoked ≥20 cigarettes/day, they were 1.41 (95 percent CI: 0.73, 2.74). Compared with those for women whose husbands did not smoke, the adjusted odds of early pregnancy loss in all conception cycles for women whose husbands smoked <20 cigarettes/day were 1.04 (95 percent CI: 0.67, 1.63) and, for women whose husbands smoked ≥20 cigarettes/day, they were 1.81 (95 percent CI: 1.00, 3.29).

Table 5 shows the relative odds of total pregnancy loss (early pregnancy loss or clinical spontaneous abortion) by husband’s smoking status in the first conception, the first two conceptions, the first three conceptions, and all conceptions. Compared with that for women whose husbands did not smoke, the adjusted odds ratio of total pregnancy loss in the first conception cycle for women whose husbands smoked <20 cigarettes/day was 0.90 (95 percent CI: 0.58, 1.41) and, for women whose husbands smoked ≥20 cigarettes/day, it was 1.17 (95 percent CI: 0.63, 2.18). Compared with that for women whose husbands did not smoke, the adjusted odds ratio of total pregnancy loss in all conception cycles for women whose husbands smoked <20 cigarettes/day was 1.01 (95 percent CI: 0.68, 1.50) and, for women whose husbands smoked ≥20 cigarettes/day, it was 1.45 (95 percent CI: 0.82, 2.56).

Table 6 shows the relative hazards of conception and clinical pregnancy among the three smoking groups. Compared with that for women whose husbands did not smoke, the adjusted hazard ratio of conception for women whose husbands smoked <20 cigarettes/day was 0.90 (95 percent CI: 0.70, 1.18) and, for women whose husbands smoked ≥20 cigarettes/day, it was 0.96 (95 percent CI: 0.66, 1.39). Compared with that for women whose husbands did not smoke, the adjusted hazard ratio of clinical pregnancy for women whose husbands smoked <20 cigarettes/day was 0.93 (95 percent CI: 0.72, 1.20) and, for women whose husbands smoked ≥20 cigarettes/day, it was 0.78 (95 percent CI: 0.55, 1.12).

DISCUSSION

Previous studies of the effect of paternal smoking on pregnancy losses were limited to clinical spontaneous abortions and did not examine early pregnancy losses, while the latter account for two thirds of all pregnancy losses (15). We found that, in a prospective cohort of women who did not smoke, women whose husbands smoked heavily (≥20 cigarettes/day) had increased risk of early pregnancy loss after any conception and had a slightly longer time to clinical pregnancy than did women whose husbands smoked less or did not smoke.

Our data suggested that excess early pregnancy loss among the group whose husbands smoked heavily might have comprised both pregnancies that in the absence of exposure would have ended in clinical spontaneous abortion and those that would not have ended in pregnancy loss. As shown in table 3, while the prevalence of clinical spontaneous abortion was consistently lowest in each consecutive conception among the group whose husbands smoked heavily (suggesting that some pregnancies that ended in early pregnancy loss would have ended in clinical spontaneous abortion otherwise), the prevalence of total pregnancy losses (early pregnancy loss plus clinical spontaneous abortion) was consistently highest in each consecutive conception among the group whose husbands smoked heavily (suggesting that some pregnancies that ended in early pregnancy loss would not have ended in pregnancy loss otherwise).

This study had several advantages. First, it was a large study in which we prospectively collected information on smoking as well as other covariates. Second, the study subjects were a homogenous cohort of young, nulliparous women who neither smoked nor drank alcohol or coffee. Most notably, we recruited women soon after marriage who planned to become pregnant over the course of the study. The subjects began to record daily diaries and to collect urine samples immediately after they stopped contraception. Furthermore, this study used a highly sensitive and specific hCG assay to detect early pregnancy loss.

The biologic mechanism through which paternal smoking might affect early pregnancy loss is not well understood but might operate through paternal, as well as maternal, pathways. Paternal smoking might increase the risk of early pregnancy loss in a small number of conceptions by inducing chromosomal damage in sperm (2325). Paternal smoking might also increase the risk of early pregnancy loss through women’s exposure to passive smoke. Although the results of previous studies have been inconsistent, exposure to maternal active smoking might affect levels of reproductive hormones such as progesterone and estrogen, which are important for conception and maintenance of pregnancy (2630). In our cohort of nonsmoking women, we found that, compared with women with nonsmoking husbands, those whose husbands smoked had significantly decreased concentrations of urinary estrone conjugates (unpublished data). Other potential passive smoke mechanisms are essentially the same as those for active smoking (2, 31), including vasocontriction and reduced placental blood flow due to nicotine (32), maternal and fetal hypoxia due to carboxyhemoglobin formation, and genotoxicity. In this study, we cannot distinguish whether the observed effect of paternal smoking on early pregnancy loss was mediated through exposure of women to passive smoke from the husband and/or through direct smoking effects on the men’s sperm.

Readers should consider several limitations of this study when interpreting the results. First, some misclassification of the magnitude of exposures was possible in this study. We based our estimation of exposures on self-reports by men of whether they currently smoked and, if so, how many cigarettes they smoked each day. Another limitation of our study was that, although the passive smoke exposure of women in our study should have been highly predicted by the amount smoked by their husbands (3), we did not measure the patterns of exposure to husbands’ smoking or exposure to sources of passive smoke other than the husband’s. We also were unable to determine whether the smoking status of the men changed during the follow-up period. We did not measure a biomarker of tobacco smoke exposure, such as urinary cotinine or hair nicotine. We were able to observe early pregnancy loss only when detectable levels of hCG were produced and excreted in a woman’s urine. The true prevalence of early pregnancy loss might have been higher than we observed. Furthermore, because the women in this study were young, nulliparous, and neither smoked nor drank alcohol or coffee, our results might not be appropriately generalized to other populations.

Another potential source of bias in our results is imperfect sensitivity and specificity of detecting early pregnancy loss, which have been described previously (33). If subjects in different exposure groups have different mean probabilities of conception in any cycle, as groups they will also have different ratios of conception cycles to nonconception cycles. Therefore, even if the sensitivity and specificity of detecting early pregnancy loss are the same for both groups, the relative odds of early pregnancy loss in one group compared with the other can be biased. We observed that women with husbands who smoked had a slightly lower mean probability of conception in each cycle compared with women who had husbands who did not smoke. We cannot precisely determine the sensitivity and specificity of our hCG assay in combination with our Bayesian analytical method to detect early pregnancy loss. We did not empirically observe anything that would have suggested less than perfect sensitivity and specificity. However, the specificity of our Bayesian analytical method could possibly be reduced by spurious high hCG values or high hCG variability in general. To determine what bias might have occurred in our results, we calculated what the true odds ratios of early pregnancy loss would have been, assuming different values of sensitivity and specificity of detecting early pregnancy loss. Generally, lower sensitivity of detecting early pregnancy loss would have resulted in bias away from the null. Lower specificity of detecting early pregnancy loss would have resulted in bias toward the null if sensitivity was good and in little bias if sensitivity was poor.

In conclusion, this prospective study links heavy paternal smoking to increased risk of early pregnancy loss. It adds to an increasing body of literature on the adverse reproductive health effects of both maternal and paternal smoking and underscores the need to target public health interventions toward both maternal and paternal smoking to improve reproductive outcomes.

ACKNOWLEDGMENTS

The study was supported in part by grant HD32505 from the National Institute of Child Health and Human Development; grants ES8337, ES-00002, ES06198, ES11682, and ES08957 from the National Institute of Environmental Health Sciences; and grants 20-FY98-0701 and 20-FY02-56 from the March of Dimes Birth Defects Foundation.

Correspondence to Dr. Xiping Xu, Department of Environmental Health, Harvard School of Public Health, 655 Huntington Avenue, FXB-1, Boston, MA 02115-6096 (e-mail: xu@hsph.harvard.edu).

TABLE 1.

Characteristics of 526 nonsmoking, nulliparous female textile workers and their husbands in Anhui, China, by husband’s smoking amount, 1996–1998

 Wives (nonsmokers)  Husbands 
 Husbands not current smokers  Husbands smoke <20 cigarettes/day  Husbands smoke ≥20 cigarettes/day  Not current smokers  Smoke <20 cigarettes/day  Smoke ≥20 cigarettes/day 
 (n = 216)   (n = 239)   (n = 71)   (n = 216)   (n = 239)   (n = 71)  
 No.  No.  No.  No.  No.  No. 
Age (years)                  
20–24 129 60  145 61  44 62  13  25 11  11 
25–29 84 39  89 37  26 37  185 86  196 82  51 72 
30–34    14  16  11 16 
35–39      
Education                  
Elementary      
Middle school 132 61  157 66  44 62  46 21  90 38  30 42 
High school 76 35  78 33  27 38  130 60  122 51  32 45 
College or above    40 19  26 11  10 
Dust exposure 207 96  228 95  70 99  58 27  83 35  15 21 
Noise exposure 190 88  215 90  68 96  76 35  96 40  22 31 
Toxin exposure    17  29 12  
Perceived stress of living                  
None or light 126 58  139 58  46 65  136 63  155 65  52 73 
Moderate 86 40  92 39  21 30  76 35  76 32  15 21 
High      
Tea drinking 99 46  110 46  38 54  112 52  142 59  52 73 
Alcohol consumption    52 24  61 26  23 32 
 Mean SD*  Mean SD  Mean SD  Mean SD  Mean SD  Mean SD 
Height (m) 1.58 0.04  1.58  0.05  1.58  0.05  1.69  0.05  1.70  0.05  1.70  0.05 
Weight (kg) 49.2  5.5  49.4  6.0  49.1  5.2  61.6  8.7  60.4  8.3  60.5  8.0 
Body mass index (kg/m219.8  1.8  19.9 2.2  19.6  1.9  21.5  2.7  20.9  2.6  20.9  2.6 
 Wives (nonsmokers)  Husbands 
 Husbands not current smokers  Husbands smoke <20 cigarettes/day  Husbands smoke ≥20 cigarettes/day  Not current smokers  Smoke <20 cigarettes/day  Smoke ≥20 cigarettes/day 
 (n = 216)   (n = 239)   (n = 71)   (n = 216)   (n = 239)   (n = 71)  
 No.  No.  No.  No.  No.  No. 
Age (years)                  
20–24 129 60  145 61  44 62  13  25 11  11 
25–29 84 39  89 37  26 37  185 86  196 82  51 72 
30–34    14  16  11 16 
35–39      
Education                  
Elementary      
Middle school 132 61  157 66  44 62  46 21  90 38  30 42 
High school 76 35  78 33  27 38  130 60  122 51  32 45 
College or above    40 19  26 11  10 
Dust exposure 207 96  228 95  70 99  58 27  83 35  15 21 
Noise exposure 190 88  215 90  68 96  76 35  96 40  22 31 
Toxin exposure    17  29 12  
Perceived stress of living                  
None or light 126 58  139 58  46 65  136 63  155 65  52 73 
Moderate 86 40  92 39  21 30  76 35  76 32  15 21 
High      
Tea drinking 99 46  110 46  38 54  112 52  142 59  52 73 
Alcohol consumption    52 24  61 26  23 32 
 Mean SD*  Mean SD  Mean SD  Mean SD  Mean SD  Mean SD 
Height (m) 1.58 0.04  1.58  0.05  1.58  0.05  1.69  0.05  1.70  0.05  1.70  0.05 
Weight (kg) 49.2  5.5  49.4  6.0  49.1  5.2  61.6  8.7  60.4  8.3  60.5  8.0 
Body mass index (kg/m219.8  1.8  19.9 2.2  19.6  1.9  21.5  2.7  20.9  2.6  20.9  2.6 

* SD, standard deviation.

TABLE 2.

Outcomes of prospective observation and prevalence of early pregnancy losses in Anhui, China, by husband’s smoking amount, 1996–1998

 Husband’s daily smoking amount 
 None currently   <20 cigarettes/day   ≥20 cigarettes/day  
 No.  No.  No. 
Outcome of prospective observation 216   239   71  
Clinical pregnancy         
No clinical spontaneous abortion 172 80  195 82  59 83 
Clinical spontaneous abortion 21 10  23 10  
Conceived but no clinical pregnancy 11   
Never conceived 12  14  
No. of early pregnancy losses by outcome of prospective observation         
No clinical spontaneous abortion 172   195   59  
0 early pregnancy loss 144 84  157  81  45  76 
1 early pregnancy loss 24 14  24  12  9  15 
2 early pregnancy losses  11   2  
≥3 early pregnancy losses   3  
Clinical spontaneous abortion 21   23    
0 early pregnancy loss 16 76  22 96  75 
1 early pregnancy loss 19   
2 early pregnancy losses   25 
≥3 early pregnancy losses   
Conceived but no clinical pregnancy 11      
0 early pregnancy loss   
1 early pregnancy loss 73  71  60 
2 early pregnancy losses 27  14  
≥3 early pregnancy losses  14  40 
 Husband’s daily smoking amount 
 None currently   <20 cigarettes/day   ≥20 cigarettes/day  
 No.  No.  No. 
Outcome of prospective observation 216   239   71  
Clinical pregnancy         
No clinical spontaneous abortion 172 80  195 82  59 83 
Clinical spontaneous abortion 21 10  23 10  
Conceived but no clinical pregnancy 11   
Never conceived 12  14  
No. of early pregnancy losses by outcome of prospective observation         
No clinical spontaneous abortion 172   195   59  
0 early pregnancy loss 144 84  157  81  45  76 
1 early pregnancy loss 24 14  24  12  9  15 
2 early pregnancy losses  11   2  
≥3 early pregnancy losses   3  
Clinical spontaneous abortion 21   23    
0 early pregnancy loss 16 76  22 96  75 
1 early pregnancy loss 19   
2 early pregnancy losses   25 
≥3 early pregnancy losses   
Conceived but no clinical pregnancy 11      
0 early pregnancy loss   
1 early pregnancy loss 73  71  60 
2 early pregnancy losses 27  14  
≥3 early pregnancy losses  14  40 
TABLE 3.

Prevalence of pregnancy losses in first, second, third, and all conceptions in Anhui, China, by husband’s smoking amount, 1996–1998

 Husband’s daily smoking amount 
 None currently   <20 cigarettes/day   ≥20 cigarettes/day  
 No.  No.  No. 
First conception 204   225   68  
Early pregnancy loss 44 22  46 20  20 29 
Clinical spontaneous abortion 16  22 10  
Total pregnancy losses 60 29  68 30  23 34 
Second conception 36   41   17  
Early pregnancy loss 22  16 39  47 
Clinical spontaneous abortion 11   
Total pregnancy losses 12 33  17 41  47 
Third conception   15    
Early pregnancy loss  27  63 
Clinical spontaneous abortion 20   13 
Total pregnancy losses 20  27  75 
All conceptions 245   288   100  
Early pregnancy loss 52 21  70 24  37 37 
Clinical spontaneous abortion 21  23  
Total pregnancy losses 73 30  93 32  41 41 
 Husband’s daily smoking amount 
 None currently   <20 cigarettes/day   ≥20 cigarettes/day  
 No.  No.  No. 
First conception 204   225   68  
Early pregnancy loss 44 22  46 20  20 29 
Clinical spontaneous abortion 16  22 10  
Total pregnancy losses 60 29  68 30  23 34 
Second conception 36   41   17  
Early pregnancy loss 22  16 39  47 
Clinical spontaneous abortion 11   
Total pregnancy losses 12 33  17 41  47 
Third conception   15    
Early pregnancy loss  27  63 
Clinical spontaneous abortion 20   13 
Total pregnancy losses 20  27  75 
All conceptions 245   288   100  
Early pregnancy loss 52 21  70 24  37 37 
Clinical spontaneous abortion 21  23  
Total pregnancy losses 73 30  93 32  41 41 
TABLE 4.

Relative odds of early pregnancy loss by husband’s smoking amount for first, first two, first three, and all conceptions in Anhui, China, 1996–1998

 No. of wives No. of conceptions Crude  Adjusted* 
   Odds ratio of early pregnancy loss 95% confidence interval Two-sided p value  Odds ratio of early pregnancy loss 95% confidence interval Two-sided p value 
First conception          
Husband not current smoker 204 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 225 0.93  0.59, 1.49 0.775  0.81  0.49, 1.33 0.400 
Husband smokes ≥20 cigarettes/day 68 68 1.52  0.82, 2.81 0.188  1.41  0.73, 2.74 0.304 
First and second conceptions†          
Husband not current smoker 204 240 Referent    Referent   
Husband smokes <20 cigarettes/day 225 266 1.10  0.71, 1.70 0.674  0.97  0.62, 1.52 0.899 
Husband smokes ≥20 cigarettes/day 68 85 1.78  1.00, 3.17 0.052  1.55  0.85, 2.81 0.153 
First, second, and third conceptions†          
Husband not current smoker 204 245 Referent    Referent   
Husband smokes <20 cigarettes/day 225 281 1.14  0.75, 1.74 0.547  1.00  0.65, 1.54 0.990 
Husband smokes ≥20 cigarettes/day 68 93 2.04  1.14, 3.65 0.016  1.73  0.95, 3.13 0.073 
All conceptions†          
Husband not current smoker 204 245 Referent    Referent   
Husband smokes <20 cigarettes/day 225 288 1.19  0.77, 1.84 0.429  1.04  0.67, 1.63 0.854 
Husband smokes ≥20 cigarettes/day 68 100 2.18  1.18, 4.02 0.013  1.81  1.00, 3.29 0.049 
 No. of wives No. of conceptions Crude  Adjusted* 
   Odds ratio of early pregnancy loss 95% confidence interval Two-sided p value  Odds ratio of early pregnancy loss 95% confidence interval Two-sided p value 
First conception          
Husband not current smoker 204 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 225 0.93  0.59, 1.49 0.775  0.81  0.49, 1.33 0.400 
Husband smokes ≥20 cigarettes/day 68 68 1.52  0.82, 2.81 0.188  1.41  0.73, 2.74 0.304 
First and second conceptions†          
Husband not current smoker 204 240 Referent    Referent   
Husband smokes <20 cigarettes/day 225 266 1.10  0.71, 1.70 0.674  0.97  0.62, 1.52 0.899 
Husband smokes ≥20 cigarettes/day 68 85 1.78  1.00, 3.17 0.052  1.55  0.85, 2.81 0.153 
First, second, and third conceptions†          
Husband not current smoker 204 245 Referent    Referent   
Husband smokes <20 cigarettes/day 225 281 1.14  0.75, 1.74 0.547  1.00  0.65, 1.54 0.990 
Husband smokes ≥20 cigarettes/day 68 93 2.04  1.14, 3.65 0.016  1.73  0.95, 3.13 0.073 
All conceptions†          
Husband not current smoker 204 245 Referent    Referent   
Husband smokes <20 cigarettes/day 225 288 1.19  0.77, 1.84 0.429  1.04  0.67, 1.63 0.854 
Husband smokes ≥20 cigarettes/day 68 100 2.18  1.18, 4.02 0.013  1.81  1.00, 3.29 0.049 

* Models adjusted for both wife’s and husband’s ages, education, perceived life stress, and exposures to dust and noise; husband’s use of alcohol, previous smoking, and exposure to toxins; and wife’s body mass index and tea drinking.

† Standard errors for both crude and adjusted models estimated to accommodate correlations in pregnancy losses among conceptions from the same woman.

TABLE 5.

Relative odds of total pregnancy loss (early pregnancy loss or clinical spontaneous abortion) by husband’s smoking amount for first, first two, first three, and all conceptions in Anhui, China, 1996–1998

 No. of wives No. of conceptions Crude  Adjusted* 
   Odds ratio of total pregnancy loss 95% confidence interval Two-sided p value  Odds ratio of total pregnancy loss 95% confidence interval Two-sided p value 
First conception          
Husband nonsmoker 204 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 225 1.04  0.69, 1.57 0.855  0.90  0.58, 1.41 0.656 
Husband smokes ≥20 cigarettes/day 68 68 1.23  0.68, 2.20 0.494  1.17  0.63, 2.18 0.614 
First and second conceptions†           
Husband nonsmoker 204 240 Referent    Referent   
Husband smokes <20 cigarettes/day 225 266 1.22  0.74, 1.62 0.645  0.99  0.66, 1.47 0.948 
Husband smokes ≥20 cigarettes/day 68 85 1.34  0.78, 2.31 0.293  1.22  0.69, 2.13 0.496 
First, second, and third conceptions†          
Husband nonsmoker 204 245 Referent    Referent   
Husband smokes <20 cigarettes/day 225 281 1.09  0.75, 1.59 0.647  0.98  0.66, 1.45 0.919 
Husband smokes ≥20 cigarettes/day 68 93 1.56  0.89, 2.72 0.120  1.39  0.78, 2.47 0.266 
All conceptions†          
Husband nonsmoker 204 245 Reference    Reference   
Husband smokes <20 cigarettes/day 225 288 1.12  0.77, 1.65 0.550  1.01  0.68, 1.50 0.964 
Husband smokes ≥20 cigarettes/day 68 100 1.64 0.92, 2.93 0.097  1.45  0.82, 2.56 0.206 
 No. of wives No. of conceptions Crude  Adjusted* 
   Odds ratio of total pregnancy loss 95% confidence interval Two-sided p value  Odds ratio of total pregnancy loss 95% confidence interval Two-sided p value 
First conception          
Husband nonsmoker 204 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 225 1.04  0.69, 1.57 0.855  0.90  0.58, 1.41 0.656 
Husband smokes ≥20 cigarettes/day 68 68 1.23  0.68, 2.20 0.494  1.17  0.63, 2.18 0.614 
First and second conceptions†           
Husband nonsmoker 204 240 Referent    Referent   
Husband smokes <20 cigarettes/day 225 266 1.22  0.74, 1.62 0.645  0.99  0.66, 1.47 0.948 
Husband smokes ≥20 cigarettes/day 68 85 1.34  0.78, 2.31 0.293  1.22  0.69, 2.13 0.496 
First, second, and third conceptions†          
Husband nonsmoker 204 245 Referent    Referent   
Husband smokes <20 cigarettes/day 225 281 1.09  0.75, 1.59 0.647  0.98  0.66, 1.45 0.919 
Husband smokes ≥20 cigarettes/day 68 93 1.56  0.89, 2.72 0.120  1.39  0.78, 2.47 0.266 
All conceptions†          
Husband nonsmoker 204 245 Reference    Reference   
Husband smokes <20 cigarettes/day 225 288 1.12  0.77, 1.65 0.550  1.01  0.68, 1.50 0.964 
Husband smokes ≥20 cigarettes/day 68 100 1.64 0.92, 2.93 0.097  1.45  0.82, 2.56 0.206 

* Models adjusted for both wife’s and husband’s ages, education, perceived life stress, and exposures to dust and noise; husband’s use of alcohol, previous smoking, and exposure to toxins; and wife’s body mass index and tea drinking.

† Standard errors for both crude and adjusted models estimated to accommodate correlations in pregnancy losses among conceptions from the same woman.

TABLE 6.

Relative hazards of conception and clinical pregnancy by husband’s smoking amount for nonsmoking, nulliparous women who conceived at least once during prospective observation in Anhui, China, 1996–1998

 No. of wives Crude  Adjusted* 
  Hazard ratio 95% confidence interval Two-sided p value  Hazard ratio 95% confidence interval Two-sided p value 
Conception         
Husband nonsmoker 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 0.91  0.71, 1.16 0.451  0.90  0.70, 1.18 0.452 
Husband smokes ≥20 cigarettes/day 68 0.93  0.65, 1.33 0.689  0.96  0.66, 1.39 0.813 
Clinical pregnancy         
Husband nonsmoker 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 0.95  0.75, 1.20 0.665  0.93  0.72, 1.20 0.587 
Husband smokes ≥20 cigarettes/day 68 0.77  0.55, 1.08 0.125  0.78  0.55, 1.12 0.180 
 No. of wives Crude  Adjusted* 
  Hazard ratio 95% confidence interval Two-sided p value  Hazard ratio 95% confidence interval Two-sided p value 
Conception         
Husband nonsmoker 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 0.91  0.71, 1.16 0.451  0.90  0.70, 1.18 0.452 
Husband smokes ≥20 cigarettes/day 68 0.93  0.65, 1.33 0.689  0.96  0.66, 1.39 0.813 
Clinical pregnancy         
Husband nonsmoker 204 Referent    Referent   
Husband smokes <20 cigarettes/day 225 0.95  0.75, 1.20 0.665  0.93  0.72, 1.20 0.587 
Husband smokes ≥20 cigarettes/day 68 0.77  0.55, 1.08 0.125  0.78  0.55, 1.12 0.180 

* Models adjusted for both wife’s and husband’s ages, education, perceived life stress, and exposures to dust and noise; husband’s use of alcohol, previous smoking, and exposure to toxins; and wife’s body mass index and tea drinking.

References

1.
Chinese Academy of Preventive Medicine. Smoking and health in China: 1996 national prevalence survey of smoking patterns. Beijing, People’s Republic of China: China Science and Technology Press, 1997.
2.
Windham GC, Eaton A, Hopkins B. Evidence for an association between environmental tobacco smoke exposure and birthweight: a meta-analysis and new data.
Pediatr Perinat Epidemiol
 
1999
;
13
:
35
–57.
3.
Sandler DP, Helsing KJ, Comstock GW, et al. Factors associated with past household exposure to tobacco smoke.
Am J Epidemiol
 
1989
;
129
:
380
–7.
4.
Chatenoud L, Parazzini F, di Cintio E, et al. Paternal and maternal smoking habits before conception and during the first trimester: relation to spontaneous abortion.
Ann Epidemiol
 
1998
;
8
:
520
–6.
5.
Himmelberger DU, Brown BW Jr, Cohen EN. Cigarette smoking during pregnancy and the occurrence of spontaneous abortion and congenital abnormality.
Am J Epidemiol
 
1978
;
108
:
470
–9.
6.
Kline J, Stein ZA, Susser M, et al. Smoking: a risk factor for spontaneous abortion.
N Engl J Med
 
1977
;
297
:
793
–6.
7.
Windham GC, Von Behren J, Waller K, et al. Exposure to environmental and mainstream tobacco smoke and risk of spontaneous abortion.
Am J Epidemiol
 
1999
;
149
:
243
–7.
8.
Windham GC, Swan SH, Fenster L. Parental cigarette smoking and the risk of spontaneous abortion.
Am J Epidemiol
 
1992
;
135
:
1394
–403.
9.
Vine MF. Smoking and male reproduction: a review.
Int J Androl
 
1996
;
19
:
323
–37.
10.
Ahlborg G Jr, Bodin L. Tobacco smoke exposure and pregnancy outcome among working women. A prospective study at prenatal care centers in Orebro County, Sweden.
Am J Epidemiol
 
1991
;
133
:
338
–47.
11.
Samet JM. New effects of active and passive smoking on reproduction?
Am J Epidemiol
 
1991
;
133
:
348
–50.
12.
Ahlborg G Jr, Bodin L. Re: “Tobacco smoke exposure and pregnancy outcome among working women: a prospective study at prenatal care centers in Orebro County, Sweden.” (Letter).
Am J Epidemiol
 
1992
;
135
:
838
.
13.
Wilcox AJ, Weinberg CR, Wehmann RE, et al. Measuring early pregnancy loss: laboratory and field methods.
Fertil Steril
 
1985
;
44
:
366
–74.
14.
Wilcox AJ, Weinberg CR, O’Connor JF, et al. Incidence of early loss of pregnancy.
N Engl J Med
 
1988
;
319
:
189
–94.
15.
Wang X, Chen C, Wang L, et al. Conception, early pregnancy loss, and time to clinical pregnancy: a population-based prospective study.
Fertil Steril
 
2003
;
70
:
577
–84.
16.
O’Connor JF, Schlatterer JP, Birken S, et al. Development of highly sensitive immunoassays to measure human chorionic gonadotropin, its beta-subunit, and beta core fragment in the urine: application to malignancies.
Cancer Res
 
1988
;
48
:
1361
–6.
17.
Bakerman S. Bakerman’s ABC’s of interpretive laboratory data. 3rd ed. Myrtle Beach, SC: Interpretive Laboratory Data, Inc, 1994.
18.
Husdan H, Rapoport A. Estimation of creatinine by the Jaffe reaction. A comparison of three methods.
Clin Chem
 
1968
;
14
:
222
–38.
19.
Gelman A, Carlin J, Stern H, et al. Bayesian data analysis. London, United Kingdom: Chapman and Hall, 1997.
20.
Dunson D. A Bayesian model of fecundability and sterility.
J Am Stat Assoc
 
2000
;
95
:
1054
–62.
21.
Liang KY, Zeger S. Longitudinal data analysis using generalized linear models.
Biometrica
 
1986
;
73
:
13
–22.
22.
Cox D. Regression models and life tables (with discussion).
J R Stat Soc (B)
 
1972
;
34
:
187
–220.
23.
Jacobs PA. The chromosome complement of human gametes.
Oxf Rev Reprod Biol
 
1992
;
14
:
47
–72.
24.
Jacobs PA, Hassold TJ. The origin of numerical chromosome abnormalities.
Adv Genet
 
1995
;
33
:
101
–33.
25.
Egozcue S, Blanco J, Vendrell JM, et al. Human male infertility: chromosome anomalies, meiotic disorders, abnormal spermatozoa and recurrent abortion.
Hum Reprod Update
 
2000
;
6
:
93
–105.
26.
Thomas EJ, Edridge W, Weddell A, et al. The impact of cigarette smoking on the plasma concentrations of gonadotrophins, ovarian steroids and androgens and upon the metabolism of oestrogens in the human female.
Hum Reprod
 
1993
;
8
:
1187
–93.
27.
Michnovicz JJ, Naganuma H, Hershcopf RJ, et al. Increased urinary catechol estrogen excretion in female smokers.
Steroids
 
1988
;
52
:
69
–83.
28.
Michnovicz JJ, Hershcopf RJ, Naganuma H, et al. Increased 2-hydroxylation of estradiol as a possible mechanism for the anti-estrogenic effect of cigarette smoking.
N Engl J Med
 
1986
;
315
:
1305
–9.
29.
Mochizuki M, Maruo T, Masuko K, et al. Effects of smoking on fetoplacental-maternal system during pregnancy.
Am J Obstet Gynecol
 
1984
;
149
:
413
–20.
30.
MacMahon B, Trichopoulos D, Cole P, et al. Cigarette smoking and urinary estrogens.
N Engl J Med
 
1982
;
307
:
1062
–5.
31.
Windham GC, Hopkins B, Fenster L, et al. Prenatal active or passive tobacco smoke exposure and the risk of preterm delivery or low birth weight.
Epidemiology
 
2000
;
11
:
427
–33.
32.
Jaakkola JJ, Jaakkola N, Zahlsen K. Fetal growth and length of gestation in relation to prenatal exposure to environmental tobacco smoke assessed by hair nicotine concentration.
Environ Health Perspect
 
2001
;
109
:
557
–61.
33.
Weinberg CR, Hertz-Picciotto I, Baird DD, et al. Efficiency and bias in studies of early pregnancy loss.
Epidemiology
 
1992
;
3
:
17
–22.