Patterns of tobacco exposure before and during pregnancy


Marlene Anderka, Massachusetts Department of Public Health, Massachusetts Center for Birth Defects Research and Prevention, 250 Washington Street, 5th floor, Boston, MA, 02108, USA. E-mail:


Objectives. To describe maternal exposure to tobacco in the three months before conception and throughout pregnancy, examine risk factors associated with tobacco exposure in pregnancy and smoking cessation, assess use of pharmacotherapy for smoking cessation and evaluate birth outcomes by smoking status. Design. A cohort of women from a multi-site United States study were asked retrospectively about their exposure to tobacco. Population. The study population was comprised of 4,667 mothers of non-malformed control infants who participated in the National Birth Defects Prevention Study from 1997 to 2003. Methods. Using computer-assisted telephone interview responses from this population-based sample, we assessed patterns of maternal smoking and exposure to environmental tobacco smoke (ETS) as well as use of pharmacotherapy for quitting smoking during pregnancy. Results. Overall, 961 (20.6%) mothers reported any smoking and 1,401 (30.0%) reported any exposure to ETS at home or work during the three months before conception through pregnancy. Of the 961 smokers, 512 (53.3%) reportedly quit smoking before or during pregnancy, including 379 (74% of quitters) in the first trimester, and 420 (43.7%) continued to smoke throughout the pregnancy. Only 2.1% of smokers reportedly used pharmacotherapy to quit smoking anytime from three months before conception through pregnancy. Low birthweight and preterm delivery rates were lowest among offspring of non-smokers and highest in offspring of those who continued to smoke throughout pregnancy. Conclusions. About one-half of mothers who reported preconceptional smoking quit before or during pregnancy. Use of pharmacotherapy to quit smoking during pregnancy was not common.


Tobacco exposure to the embryo and fetus can occur if mother smokes or is exposed to environmental tobacco smoke (ETS). Tobacco exposure during pregnancy has been associated with adverse pregnancy outcomes including preterm delivery, intrauterine growth restriction, low birthweight, perinatal deaths, and sudden infant death syndrome (1–4), as well as certain birth defects (5–12). Other outcomes such as reduced head circumference, altered brainstem development and lung structure also have been related to tobacco exposure during pregnancy (13–15). Dose-dependent associations have been found between maternal smoking and several adverse outcomes (5, 6, 8, 9, 12, 15–17). Additional evidence suggests that smoking in conjunction with some genetic factors increases the risk for adverse pregnancy outcomes (18–22). A recent review also concluded that maternal smoking might be linked to early onset smoking in offspring (16). Because maternal smoking has been associated with other high-risk behaviors including alcohol and drug use (23), lack of periconceptional multivitamins (23), and unwanted pregnancy, which has also been associated with poor outcomes (24,25), evaluation of the independent effect of smoking remains challenging.

Although nicotine is only one of the numerous toxic constituents in tobacco, it is the substance that is responsible for addiction; thus, nicotine replacement therapy (NRT) is sometimes attempted in order to assist pregnant women in quitting (26). Bupropion hydrochloric acid (HCl) (Zyban™) has been used to reduce smoking among pregnant women. Zyban™ is not a nicotine replacement, but rather an antidepressant that works neurologically to reduce the urge to smoke (27). Recently released guidelines from the Agency for Healthcare Research and Quality found inadequate evidence to determine the risk and safety of pharmacotherapy for smoking cessation during pregnancy. They recommend that providers ask pregnant women about tobacco use and provide appropriate counseling for women who smoke (28).

Cigarette smoking and ETS are relatively common modifiable risk factors for adverse pregnancy outcomes. Based on data from birth certificates during 1990–2002, about 11% of women that gave birth in the United States (US) reported smoking during pregnancy, which represented a 38% decrease from 1990 (29). Despite the decline in overall prevalence, higher rates of smoking in pregnancy persist in certain subpopulations with rates for some groups having increased. For example, between 1995–1996 and 2001–2002, 15 states reported significant increases in smoking among pregnant teens (29). In addition, smoking prevalence has been found to differ by maternal education, race, and ethnicity, and income (30–32) and quitting during pregnancy has been related to maternal race/ethnicity, age, education, alcohol use, parity, household income, the presence of other smokers in the household, and amount smoked (31, 33).

The purpose of this study was to describe the patterns of tobacco exposure (maternal smoking and ETS) in the three months before conception through delivery among the mothers of non-malformed controls who participated in a large, US multi-centered study, the National Birth Defects Prevention Study (NBDPS). We also examined risk factors associated with tobacco exposure in pregnancy, use of pharmacotherapy for smoking cessation and factors associated with quitting. Finally, we evaluated selected birth outcomes by smoking status.

Material and methods

Study population

The NBDPS, a population-based case–control study, has been described in detail elsewhere (34). This study utilized data on NBDPS non-malformed controls, who were randomly selected from electronic birth certificates or randomly selected from hospitals records (proportional to the number of birth per hospital) in NBDPS study areas. Three centers used birth certificates selection (Iowa, Massachusetts, New Jersey), three centers used hospital selection (California, New York, Texas) and two centers (Arkansas, Georgia) used hospital selection initially and then switched to birth certificate selection. Eligible controls were those born on or after 1 October 1997 and on or before 31 December 2003, with birth mothers who resided in one of eight NBDPS sites at delivery (Arkansas, California, Georgia, Iowa, Massachusetts, New Jersey, New York, Texas) and who completed interviews and reported data on the first level tobacco exposure questions (N = 4,667). Excluded were women with incomplete interviews (N = 41) and those with unknown or incomplete smoking data (N = 5). Tobacco exposure information was self-reported via the NBDPS computer-assisted telephone interviews, which were conducted in English and Spanish between 6 weeks and 24 months after delivery. NBDPS study participation was 68% for the mothers included in this analysis. The NBDPS has been approved by Institutional Review Boards in all participating sites, holds a federal certificate of confidentiality and obtains informed consent from all participating women.

Tobacco exposure

Using NBDPS interview data we delineated patterns of maternal smoking as well as ETS exposure. Interview questions included any exposure, exposure timing by month for the three months before and after conception and by trimester for the second and third trimesters, the number of cigarettes smoked by the mother/day for each of the timeframes, and whether anyone smoked cigarettes in the home and/or anyone smoked cigarettes nearby at work or school for these timeframes. We determined the proportion of infants exposed to active maternal smoking only, ETS only, and both active smoking and ETS. In addition, we described quitting by gestational timing. Because the study did not ask about duration of smoking outside the perinatal period, we could not look at the effect of duration of prior smoking on patterns of smoking during pregnancy.

Characteristics associated with tobacco use

We described patterns of tobacco exposure by demographics (maternal race/ethnicity, birth place, education, age, and household income), selected behaviors (alcohol use, illicit drug use, use of periconceptional multivitamins), pregnancy-related characteristics (infant sex, nausea and vomiting of pregnancy (NVP), pregnancy intendedness, parity, and pregnancy weight gain appropriate for BMI) and study-related variables (site and time between delivery and interview).

Pharmacotherapy for smoking cessation

We derived drug exposure information from the maternal interview, which collected information on frequency, duration, timing, and indication for any prescription or nonprescription drug use and specifically inquired about use of the nicotine patch, nicotine gum, and Zyban™. Drugs were classified and coded according to the Slone Drug Dictionary of the Slone Epidemiology Center at Boston University. We described use of NRT including use of the nicotine patch and nicotine gum (including generic nicotine and brand names) for smoking cessation. We also described use of Zyban™ (bupropion HCl). Because we did not have complete information on indication for drug use and assumed that most of the use of bupropion HCl was for depression, we only included use of bupropion HCl if it was reported as ‘Zyban’.

Data analysis

Analyses first delineated the extent, type and timing of maternal smoking and ETS exposure from three months before conception through pregnancy. We conducted stratified analysis by examining tobacco exposure by categories of demographics, selected high-risk behaviors, geographic region, and pregnancy-related characteristics. We determined the extent of quitting before and during pregnancy as well as use of pharmacotherapy for smoking cessation. Characteristics associated with quitting were determined using logistic regression analyses adjusting for characteristics associated with tobacco exposure as well as amount smoked and the presence of another smoker in the household. We also assessed the association of smoking status with low birthweight and preterm delivery among infants of singleton pregnancies using logistic regression adjusted for variables determined a priori to be associated with these poorer outcomes including maternal race/ethnicity, birth place, education, and age.


Overall 1,638 (35.1%) women reported ever smoking and 961 (20.6%) reported any smoking from three months before conception through pregnancy, referred to as ‘maternal smoking’ in this paper. In addition, 1,401 (30.0%) women reported exposure to ETS at home and/or work. Reported smoking prevalence varied across NBDPS sites with a range of 14.9–27.3% reporting maternal smoking and 19.9–43.5% with ETS exposure.

Maternal smoking was highest among non-Hispanic white women while exposure to ETS was highest among non-Hispanic black women (Table 1). Maternal smoking and ETS exposure were each more likely among women who were US born, had 12 or fewer years of education, younger age, household income of less than $50,000, reported use of alcohol or illicit drugs, had no NVP, pre-pregnancy BMI less than 19.8, or pregnancy weight gain higher than Institute of Medicine (IOM) guidelines and whose pregnancy was unintended. Maternal smoking did not differ by maternal parity, use of multivitamins or infant sex, but ETS exposure was more likely for women who did not use multivitamins, who were primiparous and whose infant was male. Maternal smoking was more likely if time between birth and interview was 12 or more months and ETS exposure was more likely if the time was more than 18 months.

Table 1. Tobacco exposure by selected characteristics.
 Maternal smoking only N (%)ETS only N (%)Both N (%)Neither N (%)
  1. Percentages may not add up to 100% due to rounding.

  2. bUnable to differentiate ‘unknown’ from ‘no’.

  3. cAny use during the month before pregnancy through the first trimester.

  4. Note: BMI, body mass index; IOM, Institute of Medicine; ETS, environmental tobacco smoke; NVP, nausea and vomiting of pregnancy.

Overall341 (7.3)781 (16.7)620 (13.3)2,925 (62.7)
 Maternal race/ethnicity    
  White, non-Hispanic252 (9.1)353 (12.7)471 (17.0)1,702 (61.3)
  Black, non-Hispanic26 (4.7)160 (28.8)69 (12.4)301 (54.1)
  Hispanic52 (4.8)219 (20.3)61 (5.7)746 (69.2)
  Other11 (4.5)49 (20.1)18 (7.4)166 (68.0)
  Unknown0 (0.0)0 (0.0)1 (9.1)10 (90.9)
 Maternal birth place    
  US316 (8.4)604 (16.0)588 (15.5)2,274 (60.1)
  Non-US25 (2.8)177 (20.1)31 (3.5)647 (73.5)
  Unknown0 (0.0)0 (0.0)1 (20.0)4 (80.0)
 Maternal education (years)    
  < 127 (5.9)172 (21.8)150 (19.0)421 (53.3)
  12119 (10.0)247 (20.8)236 (19.9)586 (49.3)
  > 12175 (6.5)362 (13.5)233 (8.7)1,911 (71.3)
  Unknown0 (0.0)0 (0.0)1 (12.5)7 (87.5)
 Maternal age (years)    
  < 25126 (8.1)362 (23.3)342 (22.0)726 (46.7)
  25–34178 (7.3)355 (14.5)238 (9.7)1,684 (68.6)
  > 3537 (5.6)64 (9.8)40 (6.1)515 (78.5)
 Household income    
  < $20,00087 (6.6)295 (22.5)256 (19.5)675 (51.4)
  $20,000–$49,999117 (8.8)257 (19.3)210 (15.8)749 (56.4)
  > $50,000106 (7.2)123 (8.4)86 (5.9)1,154 (78.6)
  Unknown31 (5.6)106 (19.2)68 (12.3)347 (62.9)
 Alcohol usec    
  Yes205 (11.3)250 (13.8)345 (19.0)1,014 (55.9)
  No132 (4.7)529 (18.7)271 (9.6)1,902 (67.1)
  Unknown4 (21.1)2 (10.5)4 (21.1)9 (47.4)
 Illicit drug usec    
  Yes25 (12.0)28 (13.5)127 (61.1)28 (13.5)
  No/unknownb316 (7.1)753 (16.9)493 (11.1)2,897 (65.0)
 Multivitamin usec    
  Yes298 (7.5)644 (16.2)513 (12.9)2,525 (63.4)
  No43 (6.3)137 (19.9)107 (15.6)400 (58.2)
 Infant sex    
  Male167 (7.1)396 (16.8)345 (14.7)1,446 (61.4)
  Female174 (7.5)385 (16.7)274 (11.9)1,475 (63.9)
  Unknown0 (0.0)0 (0.0)1 (20.0)4 (80.0)
  Yes136 (7.3)347 (18.6)286 (15.3)1,098 (58.8)
  No200 (7.3)431 (15.7)324 (11.8)1,789 (65.2)
 Trimester 1 NVP    
  Yes211 (6.7)528 (16.7)391 (12.3)2,041 (64.4)
  No130 (8.7)252 (16.9)229 (15.3)883 (59.1)
  Unknown0 (0.0)1 (50.0)0 (0.0)1 (50.0)
 Unintended pregnancy    
  Intended174 (6.3)384 (13.8)276 (9.9)1,941 (69.9)
  Mistimed61 (8.4)151 (20.9)129 (17.8)383 (52.9)
  Unwanted70 (9.1)164 (21.3)152 (19.7)385 (49.9)
  Ambivalent36 (9.4)81 (21.0)60 (15.6)208 (54.0)
  Unknown0 (0.0)1 (8.3)3 (25.0)8 (66.7)
 BMI (IOM cutoffs)    
  < 19.852 (8.2)101 (16.0)117 (18.5)362 (57.3)
  19.8–26.0173 (7.0)396 (16.0)306 (12.4)1,595 (64.6)
  26.0–29.047 (8.6)94 (17.3)69 (12.7)334 (61.4)
  > 29.063 (7.6)153 (18.4)119 (14.3)497 (59.7)
  Unknown6 (3.2)37 (19.6)9 (4.8)137 (72.5)
 Weight gain within IOM guidelines    
  Below63 (6.7)188 (19.9)103 (10.9)589 (62.5)
  Within98 (6.0)255 (15.5)187 (11.4)1,100 (67.1)
  Above174 (9.3)301 (16.0)317 (16.9)1,086 (57.8)
  Unknown6 (2.9)37 (18.0)13 (6.3)150 (72.8)
Study variables
  AR42 (7.2)135 (23.0)120 (20.5)289 (49.3)
  CA53 (7.7)87 (12.7)49 (7.2)496 (72.4)
  IA51 (9.0)81 (14.3)104 (18.3)331 (58.4)
  MA46 (7.2)66 (10.3)84 (13.1)447 (69.5)
  NY39 (6.8)114 (19.8)62 (10.8)360 (62.6)
  NJ36 (7.8)69 (15.0)77 (16.7)278 (60.4)
  TX36 (6.0)129 (21.3)77 (12.7)363 (60.0)
  CDC/Atlanta38 (7.0)100 (18.3)47 (8.6)361 (66.1)
 Time to interview (months)    
  < 12253 (7.0)611 (16.9)464 (12.8)2,283 (63.2)
  12–1869 (9.2)104 (13.8)106 (14.1)474 (62.9)
  > 1819 (6.3)66 (21.9)50 (16.6)167 (55.3)
  Unknown0 (0.0)0 (0.0)0 (0.0)1 (100.0)

As depicted in Figure 1, maternal smoking decreased 51% between three months before conception and the third trimester of pregnancy (19.6–9.6%). During the same period, ETS exposure at home declined by 9.3% and at work by 11.8%. Among maternal smokers, 14% smoked only before conception, another 33% smoked from before conception into the first trimester, another 3.9% from before conception into the second trimester and 43.7% smoked from before conception into the third trimester (Table 2).

Figure 1.

Maternal smoking and ETS exposure by gestational interval.

Note: N = 4,667; ETS, environmental tobacco smoke.

Table 2. Common patterns of maternal smoking.
Maternal smoking by gestational intervalN% Of all womena% Of smokersb
PreconceptioncTrimester 1Trimester 2Trimester 3
  1. aDenominator = 4,667 women.

  2. bDenominator = 961 smokers.

  3. cAny smoking in the three months before conception.

x   1352.914.0
xx  3206.933.3
xxx 370.83.9
Other use491.05.1

Among all women reporting maternal smoking, 29.4% smoked fewer than 5 cigarettes/day, 39.9% smoked 5–14 cigarettes/day and 24.7% smoked 15–24 cigarettes /day with the remainder having smoked 25 or more cigarettes/day.

The percentage of women smoking fewer than 15 cigarettes/day increased over the course of pregnancy whereas the percentage smoking 15 or more cigarettes/day decreased. Changes in smoking behavior occurred most often in the first trimester (Figure 2). Of the women who reported smoking during the gestational interval, 21.7% indicated a decrease in the number of cigarettes smoked/day between the month before conception and month 1 of pregnancy, 34.0% between month 1 and month 2 of pregnancy, and 24.9% from month 2 to month 3 of pregnancy. Among the 961 maternal smokers, 449 (46.7%) continued to smoke into the third trimester. Of these, 14 (3.1%) reported an increase, 188 (41.9%) a decrease, and 247 (55.0%) no change in the frequency during pregnancy overall.

Figure 2.

Changes in maternal smoking by gestational interval.

Denominator is the number of smokers in the beginning of each gestational interval. Change in the frequency category not in exact amount of cigarettes smoked.

Among women who reported any smoking between three months before conception through pregnancy, 20 (2.1%) reported use of pharmacotherapy for smoking cessation including 17 (1.7%) who reported NRT and 3 (0.3%) who reported Zyban™ (bupropion HCl). Of those reporting NRT, 10 reported nicotine patch-NOS, 6 nicotine gum-NOS and 1 Commit lozenges. Only 4 (0.4%) maternal smokers used pharmacotherapy for smoking cessation in the three months before but not during pregnancy, 7 (0.7%) both before and during pregnancy and 10 (1.0%) during pregnancy only.

Overall, 512 (53.3%) maternal smokers quit smoking before or during pregnancy. About 75% (n = 379) of these women quit in the first trimester. In addition, women whose pregnancy was intended quit on average about a month earlier than women whose pregnancy was unintended.

Quitters were more likely to be Hispanic, primiparous, have 13 or more years of education, annual household income more than $50,000, report no exposure to ETS at home, have smoked fewer than 5 cigarettes/day, report using alcohol and report not using illicit drugs during the same period. They were also more likely to have a female infant and to have weight gain above IOM guidelines (see Table 3).

Table 3. Frequency and characteristics of women who quit smoking before and during pregnancy.
 Any maternal smokingQuittersbCrudecAdjustedc,d
 NN%aOR95% CIOR95% CI
  1. aPercentages may not add up to 100% due to rounding.

  2. bQuitters: Maternal smokers who stop smoking in one gestational timing and afterwards.

  3. cExclusions: unknown race/ethnicity, birth place, education, alcohol use, amount of cigarettes smoked, infant sex, parity, T1 NVP, pregnancy intendedness (N = 15).

  4. dLogistic regression was performed, conditional on site and adjusting for the other variables in the table, to estimate the odds of being a quitter during pregnancy.

  5. eP < 0.001.

  6. fP < 0.01.

  7. gP < 0.05.

  8. hLargest amount reported if different by gestational timing.

  9. iIncluded mistimed (N = 188), unwanted (N = 218), and ambivalent (N = 94) intendedness of pregnancy.

  10. jAny use during the month before pregnancy through the first trimester.

  11. Note: NPV, nausea and vomiting of pregnancy; IOM, Institute of Medicine; ETS, environmental tobacco smoke; BMI, body mass index; Ref., reference.

 Maternal race/ethnicity       
  White, non-Hispanic71434468Ref.Ref.Ref.Ref.
  Black, non-Hispanic9350101.250.81–1.931.210.69–2.13
 Maternal birth place       
  US born89046492Ref.Ref.Ref.Ref.
  Non-US born563982.111.17–3.78g0.880.43–1.81
 Maternal education       
  < 1219586170.430.30–0.60e0.540.34–0.87g
 Maternal age       
  < 25460236470.700.59–1.010.830.57–1.22
 Household income       
  < $20,000338161320.460.32–0.66e0.750.45–1.27
  > $50,00018812525Ref.Ref.Ref.Ref.
 Alcohol usej       
 Illicit drug usej       
 Exposure to ETS at home       
 Amount smokedh       
  < 5/day27619940Ref.Ref.Ref.Ref.
  > 25/day561120.100.05–0.19e0.150.07–0.33e
 Multivitamin usej       
 Infant sex       
 Trimester 1 NVP       
 Unintended pregnancy       
 BMI (IOM cutoffs)       
  < 19.816779160.710.50–1.020.780.51–1.19
  > 29.018091180.810.58–1.150.910.60–1.38
 Weight gain within IOM guideline

Among singleton births of nonsmokers, 4.2% were low birthweight, compared to 5.1% of those for quitters and 8.0% of those who continued to smoke throughout pregnancy. In logistic regression analyses adjusted for maternal age, race/ethnicity, education, and birth place, mothers who continued to smoke throughout pregnancy were more likely than nonsmokers to deliver a low birthweight infant (adjusted odds ratio, aOR, 1.89, 95% CI 1.24–2.86).

Among nonsmokers with singleton births, 7.6% delivered prematurely (< 37 weeks), compared to 8.0% of quitters and 11.5% of those who continued to smoke throughout pregnancy. In logistic regression analyses adjusted for maternal age, race/ethnicity, education, and birth place, mothers who continued to smoke throughout pregnancy were more likely than nonsmokers to deliver a premature infant (aOR, 1.59, 95% CI 1.13–2.25).


Using self-reports from a population-based sample of women from eight sites in the US, this study determined that 20.6% reported any smoking from three months before conception through pregnancy, and 30.0% reported exposure to ETS at home or work. Very few women (2.1% of smokers) reported using medications for smoking cessation during pregnancy. Of the mothers who smoked, 14% quit before conception. Overall, about one-half of smokers quit smoking before or during pregnancy, and most of these, about three quarters, did so in the first trimester. About 10% of women smoked throughout their entire pregnancy.

A previous study based on 2001 data from the Pregnancy Risk Assessment Monitoring System (PRAMS), found that 22.5% of women reported smoking before pregnancy (35). In our study, which is based on women who delivered from 1997 to 2003, 19.1% reported smoking before pregnancy. Assuming that the survey instruments in the two studies had similar sensitivity, these smoking rates from two different US samples are comparable.

In agreement with our findings, a review article reported that most women who quit in pregnancy quit in early pregnancy (36). Quit rates were slightly higher in our study than those reported in PRAMS (53.3 vs. 44.5%, respectively) (37), possibly due to differences in the populations surveyed or increased maternal quitting in more recent years. Analogous to our risk factor findings, a number of investigators have found higher maternal smoking rates among white women and women with lower education (36, 38).

Less than 3% of maternal smokers reported use of pharmacotherapy for smoking cessation anytime from three months before conception through pregnancy. Most of the NRT products for smoking cessation including some nicotine patches as well as nicotine gum and lozenges are sold over-the-counter. The non-nicotine Zyban™ and nicotine nasal spray and inhaler are available by prescription ( (39). Given that there is little information on the safety and effectiveness of these products in pregnancy and that recent US clinical recommendations do not advise use in pregnancy, pregnant women may not be comfortable using these products and/or may not have been encouraged to do so by their health care providers (28, 40). Alternatively, the NBDPS may have underestimated use of these products.

The percentage of women smoking fewer than 15 cigarettes/day increased over the course of pregnancy whereas the percent smoking 15 or more cigarettes/day decreased. Changes in smoking status and frequency occurred most often in the first trimester, which coincides with when women recognize they are pregnant or have their first prenatal visit. Women who intended to become pregnant reduced smoking or quit on average about a month earlier than other women.

Considering singleton births only, low birthweight and preterm delivery rates were lowest among the offspring of non-smokers, highest among the offspring of those who continued smoking throughout pregnancy, and intermediate among the offspring of quitters. This would support the view that, whereas not smoking is best, quitting during pregnancy may still improve pregnancy outcomes compared to continued smoking.

Strengths of this study included the availability of detailed information about the timing of maternal tobacco exposure and the ability to control for variables such as alcohol and illicit drug use that are often associated with smoking.

A number of limitations should be considered in interpreting the results of this study. These include the self-reported nature of the exposure data, the retrospective study design and the sample selection. Information on maternal smoking and ETS was obtained from retrospective maternal interviews. A previous study that used cotinine levels to validate self-reports found that self-reports of maternal smoking in both early and late pregnancy were reliable, but those for ETS exposure were less so (41). A similar prospective study using cotinine found that reports for 85% of pregnant women who reported smoking and 95% of women who denied smoking were confirmed by cotinine testing (42). Other studies, including one review article, have found that self-reports of smoking during pregnancy underestimated the true frequency of smoking (43–45). Another limitation is that this study was based on non-malformed controls in the NBDPS, who are a population-based random sample selected from birth certificates or birth hospitals. In generalizing these results readers should consider that 68% of eligible non-malformed controls participated in the NBDPS, the study sample excluded mothers whose baby had a major birth defect, and that an analysis comparing characteristics of the birthing population to the selected and interviewed controls in the NBDPS demonstrated that interviewed controls tended to be slightly older and more educated than the target population (46). Women in this study were reporting for behaviors that occurred between 1997 and 2003. It is possible that smoking behaviors have changed in more recent years.


Because about 20% of pregnant women report smoking sometime between three months before conception and delivery, and maternal smoking is detrimental to the mother and baby, it is important to determine the best approach to quitting. Characteristics associated with quitting should be considered when targeting smoking cessation interventions. One option to support cessation, NRT, is being used in pregnancy only by a small fraction of smokers but the efficacy and safety of use during pregnancy needs further study. Limiting ETS exposure at home or work is something that pregnant woman may not be able to control on their own and they may need help from the public health community to reduce ETS exposure.


This work was funded in part by Cooperative Agreement number U50/CCU113247 from the Centers for Disease Control and Prevention. We also thank Chia-ling Liu for computer programming support.

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