Volume 203, Issue 4 , Pages 334.e1-334.e8, October 2010
The risk of adverse pregnancy outcomes is increased in preeclamptic women who smoke compared with nonpreeclamptic women who do not smoke
Article Outline
Objective
Maternal smoking and preeclampsia independently increase the risk of adverse pregnancy outcomes; however, smoking decreases the risk of preeclampsia. We sought to estimate the risk of adverse pregnancy outcomes among preeclamptic women who smoke and hypothesized that this risk would be increased, compared with nonpreeclamptic women who smoke or preeclamptic women who do not smoke.
Study Design
With the use of the Niday Perinatal Database and multiple logistic regressions, we estimated the risk of adverse pregnancy outcomes in nonpreeclamptic women who smoke, preeclamptic women who do not smoke, and preeclamptic women who smoke in relation to nonpreeclamptic women who do not smoke.
Results
The incidence of adverse pregnancy outcomes was more than twice as high among preeclamptic women who smoke as among nonpreeclamptic women who do not smoke. The following data were observed: small-for-gestational-age infant (odds ratio [OR], 3.40; 95% CI, 2.27–4.89), preterm birth (OR, 5.77; 95% CI, 4.50–7.35), very preterm birth (OR, 5.44; 95% CI, 3.51–8.11), abruption (OR, 6.16; 95% CI, 3.05–11.01), Apgar <4 at 5 minutes (OR, 3.11; 95% CI, 1.48–5.72), and stillbirth (OR, 3.39; 95% CI, 1.33–6.99).
Conclusion
Smoking decreases the risk of preeclampsia, but smokers with preeclampsia have an increased risk for adverse pregnancy outcomes.
Key words: adverse pregnancy outcome, preeclampsia, smoking
The prevalence of smoking among pregnant women is a major ongoing health concern, especially considering the number of adverse pregnancy outcomes that are associated with smoking during pregnancy. The related adverse outcomes include spontaneous abortion, stillbirth, preterm birth, and fetal growth restriction.1, 2, 3 Additional concerns include the possibility of long-term adverse effects on an infant,3 such as neurodevelopmental disorders and cancers.2 Although smoking generally is associated with increased perinatal morbidity and death,1 interestingly, it has been shown to have a protective effect on the risk of the development of preeclampsia.4, 5, 6, 7 Indeed, smoking reduces the risk of preeclampsia8, 9 and decreases its incidence in a dose-effect manner;10, 11 it may “protect”11, 12, 13 against or “mask”12 its development. The risk of preeclampsia among pregnant women who smoke is 32% lower than that among nonsmoking pregnant women.4
Preeclampsia is a disease of pregnancy that is characterized by increased maternal blood pressure and proteinurea.14 It is a major cause of maternal death and morbidity,15 and like smoking, it also causes perinatal deaths, preterm birth, and intrauterine growth restriction.15 Despite the apparent benefits of smoking on the risk of preeclampsia, the biologic linkage between smoking and preeclampsia is not well understood.7 Although it has been established that the risk of preeclampsia is decreased in pregnant smokers, the consequent risks that result when preeclampsia does develop in these women have not been characterized.
We undertook a study to estimate the interaction between the effects of smoking and preeclampsia on the risk of adverse pregnancy outcomes. Our objective was to determine the risk of the development of a wide range of adverse pregnancy outcomes among pregnant smokers who experience preeclampsia in relation to nonpreeclamptic nonsmoking pregnant women. We hypothesized that the risks of adverse pregnancy outcomes would be increased among smokers who are preeclamptic, compared with pregnant smokers who are not preeclamptic or nonsmokers who are preeclamptic.
Materials and Methods
This study was based on the 2004-2006 Niday Perinatal Database, which is part of the Ontario Perinatal Surveillance System and has included >95% of births in Ontario, Canada. There are 82 participating sites that include both hospitals and midwifery groups; 359,747 births were recorded between 2004 and 2006. Sites were permitted to enter data directly into the database (64 sites) or to upload data from their own databases (18 hospitals). The database includes information on the number of women who give birth and infants who are born in the province of Ontario and includes maternal demographic information, maternal behavior, maternal health problems, maternal complications, intrapartum complications and interventions, birth outcomes, and infant health. Ethical approval was obtained from The Ottawa Hospital Research Ethics Board to analyze the information from this database for our study.
The chosen sample size was based on the entire data of the large convenience sample of the 3-year Ontario database for the years 2004-2006. For the analysis, we first compared the demographic information between the smoking and nonsmoking groups, for which smoking was defined as any smoking during pregnancy (regardless of the amount). Second, we compared the demographic information between the preeclamptic and nonpreeclamptic groups, for which preeclampsia was defined (by the Niday Perinatal Database) as the development of hypertension ([1] a rise in systolic pressure of at least 30 mm Hg, a rise in diastolic pressure of at least 15 mm Hg, or a diastolic pressure of at least 90 mm Hg; [2] a blood pressure of 140/90 mm Hg on at least 2 occasions at least 6 hours apart; [3] a mean arterial pressure of 105 mm Hg) with proteinuria (determined by either urine dipstick analysis or protein concentration >0.3 g in a 24-hour urine collection) that occurred after 20 weeks of gestation. In all cases, we used chi-squared statistical tests. Third, we compared the incidences of selected fetal outcomes that were associated with maternal smoking or preeclampsia, compared with either all nonsmoking or nonpreeclamptic pregnancies, respectively, using chi-squared statistical tests. With the use of multiple logistic regression, the adjusted odds ratios were determined. Fourth, using chi-squared statistical tests and multiple logistic regression, we estimated the effect on adverse pregnancy outcomes when smoking and preeclampsia both occurred, in comparison to nonsmoking nonpreeclamptic pregnancies. As a control, we performed the same analyses for the group of smoking nonpreeclamptic women and nonsmoking preeclamptic women. The outcomes that were assessed in this study included small-for-gestational-age births (<3rd percentile), preterm birth (delivery at <37 weeks of gestation), very preterm birth (<32 weeks of gestation), placental abruption (premature separation of a normally implanted placenta that resulted in retroplacental bleeding after week 20 of gestation and before the fetus was delivered), venous cord pH (<7.0), Apgar reading (<4 at 5 minutes), and stillbirth (death of fetus at >20 weeks of gestation). Potential confounding variables included maternal age, parity, and multiple gestation. All analyses were performed with SAS-PC statistical software (version 9.1; SAS Institute Inc, Cary, NC).
Results
Women who smoked during pregnancy tended to be young (<25 years old) and were less likely to have had a prenatal visit during their first trimester, to have used reproductive technologies, to have had a multiple pregnancy, or to have been diabetic (Table 1). Preeclampsia was more likely in women <25 and >35 years old, nulliparous women and women who had a prenatal visit during their first trimester, women who used assisted reproductive technologies, women with multiple gestation, and women who had either diabetes mellitus or high blood pressure (Table 2).
TABLE 1. Demographics of smoking
| Variable | Nonsmoker, n (%)a | Smoker, n (%)a |
|---|---|---|
| Age, y | ||
 <19 | 3838(1.40) | 2456(7.12) |
| 19-25 | 33,674(12.30) | 12,006(34.79) |
| 26-30 | 77,533(28.33) | 9547(27.67) |
| 31-35 | 98,834(36.11) | 6526(18.91) |
| >35 | 59,810(21.85) | 3973(11.51) |
| Parity | ||
| 0 | 121,889(45.19) | 15,712(45.91) |
| ≥1 | 147,865(54.81) | 18,511(54.09) |
| First prenatal visit | 108,965(83.28) | 15,805(77.43) |
| Assisted reproductive technologies | 3393(2.29) | 184(0.79) |
| Multiple gestation (twins) | 9679(3.54) | 1083(3.14) |
| Medical problems | ||
| Diabetes mellitusb | 9529(4.84) | 1078(4.08) |
| Chronic hypertension | 7649(4.00) | 1056(3.99) |
aχ2 statistical tests were used to calculate numbers and percentages; |
bRepresents insulin-dependent, noninsulin-dependent, and gestational diabetes mellitus. |
TABLE 2. Demographics of preeclampsia
| Variable | Nonpreeclampsia, n (%)a | Preeclampsia, n (%) |
|---|---|---|
| Age, y | ||
| <19 | 5124(2.15) | 103(3.03) |
| 19-25 | 36,265(15.20) | 526(15.45) |
| 26-30 | 67,072(28.11) | 915(26.88) |
| 31-35 | 80,597(33.78) | 1077(31.64) |
| >35 | 49,564(20.77) | 783(23.00) |
| Parity | ||
| 0 | 105,595(45.05) | 2234(66.33) |
| ≥1 | 128,822(54.95) | 1134(33.67) |
| First prenatal visit | 117,986(81.03) | 1305(85.24) |
| Assisted reproductive technologies | 3562(2.07) | 106(4.73) |
| Multiple gestation (twins) | 8776(3.68) | 357(10.49) |
| Medical problems | ||
| Diabetes mellitusb | 11,048(4.70) | 256(7.80) |
| Chronic hypertension | 8957(3.81) | 443(13.50) |
aχ2 statistical tests were used to calculate numbers and percentages; |
bRepresents insulin-dependent, noninsulin-dependent, and gestational diabetes mellitus. |
To consider which outcomes were associated with smoking and preeclampsia, we studied the groups of all women who smoked during pregnancy and all women who experienced preeclampsia. The incidence of adverse pregnancy outcomes that were studied was higher among smokers, compared with nonsmoking pregnant women (Table 3). One exception was the incidence of preeclampsia, which was reduced among smokers compared with nonsmokers (adjusted odds ratio, 0.83; 95% confidence interval, 0.74–0.94). Additionally, the incidences of adverse pregnancy outcomes that were studied were increased among women with preeclampsia, compared with nonpreeclamptic women (Table 4). The incidence of venous cord pH <7.0 was similar among the smoking and preeclamptic groups, compared with the nonsmoking and nonpreeclamptic groups, respectively. The incidence of stillbirth was not increased in the preeclamptic group.
TABLE 3. Adverse pregnancy outcomes associated with smoking
| Outcome | Smoker, n (%) | Nonsmoker, n (%) | Adjusted OR (95% CI)a |
|---|---|---|---|
| Small-for-gestational-age <3 | 1521(4.42) | 7038(2.58) | 1.73(1.63–1.84) |
| Preterm birth | |||
| <37 | 3741(10.87) | 22603(8.27) | 1.26(1.22–1.31) |
| <32 | 828(2.41) | 4156(1.52) | 1.63(1.51–1.77) |
| Abruption | 247(0.92) | 1007(0.52) | 1.82(1.57–2.11) |
| Preeclampsia | 324(1.21) | 2888(1.49) | 0.83(0.74–0.94) |
| pH <7.0 | 84(0.42) | 589(0.43) | 1.10(0.87–1.39) |
| Apgar <4 at 5 mins | 404(1.18) | 2144(0.79) | 1.47(1.31–1.64) |
| Admission to NICU | 1600(7.11) | 8178(5.74) | 1.31(1.24–1.39) |
| Stillbirth | 269(0.78) | 1343(0.49) | 1.58(1.38–1.81) |
aAdjusted for maternal age, parity, and multiple gestation. |
TABLE 4. Adverse pregnancy outcomes associated with preeclampsia
| Outcome | Preeclampsia, n (%)a | Nonpreeclampsia, n (%)a | Adjusted odds ratio (95% CI)b |
|---|---|---|---|
| Small-for-gestational-age <3rd percentile | 190(5.59) | 6610(2.78) | 1.62(1.39–1.88) |
| Preterm birth, wk | |||
| <37 | 1314(38.65) | 21,003(8.84) | 6.16(5.71–6.65) |
| <32 | 303(8.91) | 4141(1.74) | 4.06(3.55–4.62) |
| Abruption | 33(0.97) | 1347(0.56) | 1.60(1.09–2.25) |
| pH <7.0 | 16(0.81) | 640(0.44) | 1.67(0.97–2.66) |
| Apgar score <4 at 5 min | 59(1.74) | 2137(0.90) | 1.57(1.19–2.03) |
| Admission to neonatal intensive care unit | 428(20.91) | 9712(6.27) | 3.39(3.02–3.81) |
| Stillbirth | 31(0.91) | 1369(0.57) | 1.39(0.95–1.96) |
aχ2 statistical tests were used to calculate numbers and percentages; |
bCalculated by multiple logistic regression; adjusted for maternal age, parity, and multiple gestation. |
The independent effects of smoking and preeclampsia on adverse pregnancy outcomes were observed by a comparison of smoking nonpreeclamptic women (Table 5) and nonsmoking preeclamptic women (Table 6) to nonpreeclamptic nonsmoking women. Nonpreeclamptic smoking women had an increased risk of small-for-gestational-age infant, preterm birth, very preterm birth, abruption, Apgar score <4 at 5 minutes, and stillbirth (there were no significant differences in the incidence of pH <7.0). Similarly, nonsmoking preeclamptic women had an increased risk of small-for-gestational-age fetus, preterm birth, very preterm birth, and Apgar score <4 at 5 minutes (there were no significant differences in the incidences of abruption, pH <7.0, and stillbirth). By comparing the population of smoking preeclamptic women with the nonpreeclamptic nonsmoking group, we observed the effect on pregnancy outcomes when both exposures occur concurrently in an individual. Remarkably, the incidences of all adverse pregnancy outcomes that were studied were more than twice as great in the smoking preeclamptic group, compared with the nonsmoking nonpreeclamptic group (Table 7).
TABLE 5. Adverse pregnancy outcomes: smoker, no preeclampsia
| Outcome | Nonsmoker with no preeclampsia, n (%)a | Smoker with no preeclampsia, n (%)a | Adjusted odds ratio (95% CI)b |
|---|---|---|---|
| Small-for-gestational age <3rd percentile | 4684(2.45) | 1140(4.31) | 1.80(1.68–1.93) |
| Preterm birth, wk | |||
| <37 | 15,857(8.30) | 2828(10.68) | 1.45(1.38–1.51) |
| <32 | 2940(1.54) | 614(2.32) | 1.62(1.48–1.78) |
| Abruption | 987(0.52) | 236(0.89) | 1.80(1.54–2.08) |
| pH <7.0 | 520(0.44) | 75(0.42) | 1.08(0.84–1.38) |
| Apgar score <4 at 5 min | 1549(0.81) | 306(1.16) | 1.44(1.26–1.64) |
| Stillbirth | 982(0.51) | 203(0.76) | 1.54(1.31–1.80) |
aχ2statistical tests were used to calculate numbers and percentages; |
bCalculated by multiple logistic regression; adjusted for maternal age, parity, and multiple gestation. |
TABLE 6. Adverse pregnancy outcomes: nonsmoker, preeclampsia
| Outcome | Nonsmoker with no preeclampsia, n (%)a | Nonsmoker with preeclampsia, n (%)a | Adjusted odds ratio (95% CI)b |
|---|---|---|---|
| Small-for-gestational age <3rd percentile | 4684(2.45) | 150(5.20) | 1.63(1.37–1.93) |
| Preterm birth, wk | |||
| <37 | 15,857(8.30) | 1114(38.63) | 6.57(6.04–7.14) |
| <32 | 2940(1.54) | 254(8.81) | 4.37(3.77–5.05) |
| Abruption | 987(0.52) | 20(0.69) | 1.27(0.79–1.93) |
| pH <7.0 | 520(0.44) | 13(0.78) | 1.64(0.89–2.73) |
| Apgar score <4 at 5 min | 1549(0.81) | 47(1.63) | 1.59(1.16–2.12) |
| Stillbirth | 982(0.51) | 25(0.87) | 1.42(0.93–2.08) |
aχ2statistical tests were used to calculate numbers and percentages; |
bCalculated by multiple logistic regression; adjusted for maternal age, parity, and multiple gestation. |
TABLE 7. Adverse pregnancy outcomes: smoker, preeclampsia
| Outcome | Nonsmoker with no preeclampsia, n (%)a | Smoker with preeclampsia, n (%)a | Adjusted odds ratio (95% CI)b |
|---|---|---|---|
| Small-for-gestational age <3rd percentile | 4684(2.45) | 30(9.26) | 3.40(2.27–4.89) |
| Preterm birth, wk | |||
| <37 | 15,857(8.30) | 107(33.02) | 5.77(4.50–7.35) |
| <32 | 2940(1.54) | 28(8.64) | 5.44(3.51–8.11) |
| Abruption | 987(0.52) | 11(3.40) | 6.16(3.05–11.01) |
| pH <7.0 | 520(0.44) | 3(1.36) | 3.20(0.79–8.45) |
| Apgar score <4 at 5 min | 1549(0.81) | 9(2.80) | 3.11(1.48–5.72) |
| Stillbirth | 982(0.51) | 6(1.85) | 3.39(1.33–6.99) |
aχ2 statistical tests were used to calculate numbers and percentages; |
bCalculated by multiple logistic regression; adjusted for maternal age, parity, and multiple gestation. |
Comment
The key findings of our study indicate that both smoking and preeclampsia independently increase the risk for adverse pregnancy outcomes. Specifically, among the smoking nonpreeclamptic population, the incidence of small-for-gestational-age infants, preterm births, very preterm births, abruption, Apgar scores of <4 at 5 minutes, and stillbirth were increased significantly, compared with the nonsmoking, nonpreeclamptic group. Among nonsmoking preeclamptic women, the incidence of small-for-gestational-age infants, preterm birth, very preterm birth, and Apgar scores of <4 at 5 minutes were also increased. Smoking was not associated with a significant alteration of low venous cord pH (<7.0), nor was preeclampsia associated with a significant increase in the incidence of abruption, venous cord pH <7.0, or stillbirth. In contrast, the risk of preeclampsia was decreased in smokers, compared with nonsmokers. Smoking had a protective effect on the risk of preeclampsia. However, when women were exposed to smoking and experienced preeclampsia, an additive effect was observed in regards to the incidence of adverse pregnancy outcomes. Importantly, among smoking preeclamptic women, the incidence of adverse pregnancy outcomes was more than double that of nonsmoking, nonpreeclamptic pregnant women.
Our study has 4 major strengths. First, the study included a large sample size (359,747 individuals), which lends to our confidence in the results. Second, because of the detailed data that were included in the Niday Perinatal Database, a wide range of outcomes could be studied, and confounders could be controlled for through adjustment. These characteristics of our study resulted in very robust results and gave us confidence in our population and our analysis. Third, our study was population-based, which allowed the results to be interpreted at a population level and provided evidence for clinicians and policymakers acting at the level of primary prevention. Fourth, our dataset comes from an Ontario population in which universal healthcare is provided. Thus, our population should include a diverse group of individuals because access to the healthcare in Ontario is not limited to those in higher income brackets. The Ontario population tends to include a broad and diverse group of both rural and urban individuals and individuals with diverse ethnic backgrounds; this may allow for extrapolation of our results to populations in other Canadian populations. Our extensive observations of a diverse population may provide clinicians with a better understanding of the broad, large-scale, and serious effects of the development of preeclampsia as a smoker. We have no reason to believe that our study was affected by bias, because confounding by age, parity, and multiple gestation were controlled for with logistic regression.
There are several limitations related to this study. There may have been some administrative error involved in the data collection, such as coding error or misclassification error. However, this error is random and minimized by our large sample size. Also, all data regarding smoking habits was self-reported; therefore, because of the social stigma that is associated with smoking during pregnancy, there may have been some under-reporting. We also did not have data about the trimester of exposure or the duration of exposure. We did not control for the amount of smoking during pregnancy or environmental smoke exposure. Finally, we did not control for the demographics of maternal body mass index or ethnicity in our control and study groups.
Our findings regarding adverse pregnancy outcomes that are associated with smoking and preeclampsia are similar to findings from other reported studies. For example, the incidences of small-for-gestational-age fetus,16 preterm birth,10 and stillbirth17 have been found to be increased among smokers; our data also reveals these associations. The incidence of abruption among smokers was found to be increased with an odds ratio of 1.78 among women in the hospital-based studies of a metaanalysis,7 which is consistent with our adjusted odds ratio of 1.80. Additionally, as was found among our population, most of the literature suggests that the incidence of preeclampsia is reduced among smokers, compared with nonsmokers.4, 5, 6, 7, 8, 9, 10, 11 Venous cord pH <7.0, Apgar scores of <4 at 5 minutes after birth, and admission to neonatal intensive care (NICU) are all predictors of poor neonatal and fetal health. Although we did not observe an impact on venous cord pH, low Apgar scores and frequency of admission to NICU were both increased among smokers, which is consistent with our understanding that smoking causes increased perinatal morbidity.1 The incidences of adverse pregnancy outcomes have also been associated with preeclampsia. In particular, preeclamptic women have been shown to have an increased risk of intrauterine growth restriction with an odds ratio of 2.16,18 which is similar to our findings. Preeclampsia has also been associated, in varying degrees, to most classifications of preterm births19 and has also been associated with placental abruption.20 Fetal acidosis has been associated with preeclampsia.21 Admission to NICU because of preeclampsia has been studied, but it was suggested that more research in this area is required.22 Associations were also found between stillbirth and preeclampsia.23 These associations are consistent with our findings.
Our results are in agreement with other studies that have examined the interacting effects of smoking and preeclampsia. Although the notion of the interacting effects of smoking and preeclampsia on the risk of adverse pregnancy outcomes is not novel, our study presents new information based on its scale and its implementation among a Canadian population. A number of other studies have found a similar interaction between the effects of smoking and preeclampsia. An earlier study included a much smaller sample size (n = 2543 women), compared with our own (n = 359,747 women) and studied only 2 pregnancy outcomes: low birthweight and perinatal death.24 Similarly, another study (n = 18,631 women) investigated only perinatal death.25 A large Swedish study (n = 317,352) evaluated only a limited number of pregnancy outcomes: abruption, perinatal death, and small-for-gestational-age infants.5 In contrast to our study, the results from the Swedish study may not be so generalizable because of their much more homogenous study population. Another study of Western European pregnant white women (n = 1001) found by comparing preeclamptic women who had either smoked or never smoked or ceased smoking that, although smoking in preeclamptic pregnancies exacerbated perinatal morbidity, stopping smoking decreased the risk.26 In that study, preterm delivery, intrauterine growth retardation, admission to the NICU, and perinatal death were evaluated. However, because that study included only a specific group of women, it was suggested by the author that the findings may not be generalizable to other populations. Also, it was noted that the number of smoking pregnant women among the study group, who had all been previously diagnosed with preeclampsia, was low. Among a unique inner-city, predominantly African American population,27 the same interaction was also observed; however, these results are also less extrapolable.
Two studies present contradicting findings to our own. Rasmussen and Irgens28 found that the effects of pregnancy-induced hypertension and smoking were not “synergistic,” which suggested that they act by different mechanisms to exert their effects. This Norwegian study included a large sample size (n = 215,598), but its weaknesses are that the only outcome studied was low birthweight; confounding by parity, weight gain and prepregnancy weight was not controlled for. It was different from other studies because it examined the combined effects of smoking in the different subgroups of hypertensive complications of pregnancy. Another Norwegian study by Hogberg et al29 showed similar findings that the effects of hypertensive pregnancy disorders and smoking on abruption are “independent and additive,” again suggesting that smoking and preeclampsia exert their effects through separate mechanisms. Unlike our study, this study included much more detailed information regarding the smoking habits of women, including number of cigarettes smoked, and information about the class of hypertensive complication of pregnancy. However, under-reporting of smoking may have been an issue, and the study of abruption as the only outcome meant a sample size of only 814 cases of abruption to be studied and therefore less robust results.
Our study was conducted as an epidemiologic, population-based study that was intended to characterize the risks that are associated when smoking and preeclampsia occur simultaneously in pregnant women. Although our study does not provide the results to make any mechanistic conclusions regarding the interacting effects of smoking and preeclampsia, potential mechanisms have been elucidated. Cnattignius et al5 suggest that smoking and preeclampsia may act synergistically on fetoplacental circulation. Cigarettes contain many toxic substances with hypotensive effects (eg, thiocyanate) and nicotine, which inhibits the production of thromboxane (potent vasoconstrictor and platelet aggregation stimulator), and thus may be protective for preeclampsia. Smoking, however, also reduces prostacyclin (vasodilator) production. Additionally, vasoconstriction in preeclampsia may be mediated by reduced prostacyclin production and increased thromboxane production. This increased vascular resistance increases arterial blood pressure, can lead to placental hypoxemia, and may be compounded by the reduction in prostacyclin among smokers. Placental hypoxemia can lead to placental infarcts and necrosis, which may increase the risk of fetal hypoxia, abruption, and death. Thus, for smokers who actually have preeclampsia, there may be profound changes in fetoplacental circulation, which can lead to increased risk of adverse pregnancy outcomes. Andrews and McGarry 25 indicate that, when hypertension occurs, there already may be an element of poor fetal growth that, if associated with smoking in pregnancy, contributes to higher perinatal mortality rates. Pipkin26 suggests that smoking desensitizes the vascular endothelium so that the response to the metabolic insults preceding preeclampsia is reduced; however, when this protection is overwhelmed by more severe stimuli, more severe disease occurs. Endogenous nitric oxide (a potent vasodilator), which is reduced in both smokers and preeclamptic women, may be involved in this pathway.
The clinical implications of our study are important, because smokers who experience preeclampsia clearly warrant much higher levels of surveillance during their pregnancies and may require intervention. Our findings emphasize the need to promote smoking cessation during pregnancy and indicate a population that is at a higher risk of adverse pregnancy outcomes that should be subject to higher levels of counseling. Effective treatment is needed to improve smoking cessation rates,30, 31, 32, 33, 34 but this treatment is not always provided.34, 35 Helping clinicians and policymakers to better understand factors that put women at risk for adverse pregnancy outcomes will help them to target their efforts toward those women at highest risk. Additionally, a study by England et al36 found that women who smoke, but quit before pregnancy, do not have the reduced risk of preeclampsia as is seen in pregnant women who smoke. This underscores the need to target smokers even before pregnancy to reduce the number of smokers who could be subject to the higher risk of adverse outcomes if preeclampsia were to develop in their pregnancies. Finally, the results of our study are important because observing the epidemiologic interaction between the effects of smoking and preeclampsia on the risk of adverse pregnancy outcomes may provide new directions or insight for researchers who perform mechanistic studies regarding these risk factors.
In conclusion, we have shown from a large population database in Ontario, Canada, that, although smoking decreases the risk of preeclampsia, when smokers experience preeclampsia, the risk of adverse pregnancy outcomes is increased compared with that observed in nonpreeclamptic women who are nonsmokers. Future research in this area should include measures of urinary cotinine levels to confirm and quantify smoking habits. The risk of adverse outcomes in preeclamptic women that are associated with environmental tobacco exposure (eg, having a spouse who smokes) and the risks among preeclamptic women who quit smoking either before or during pregnancy should be studied. Future research should also be aimed at strengthening our understanding of both the factors that lead to adverse pregnancy outcomes and the policies, care, and support that are effective in the prevention of these outcomes from occurring in the first place.
Acknowledgments
We thank Ruth White, Diane Gorley, and Jim Bottomley for their important contributions to this project.
References
- . Canadian Perinatal Health Report: Canadian Perinatal Surveillance System, Health Canada. In: Ottawa, Ontario, Canada: Minister of Public Works and Government Services Canada; 2003;p. 86–87
- . The epidemiology of smoking during pregnancy: smoking prevalence, maternal characteristics and pregnancy outcomes. Nicotine Tob Res. 2004;6:S125–S140
- . Smoking during pregnancy in the 1990s. Natl Vital Stat Rep. 2001;49:1–14
- . Cigarette smoking during pregnancy and risk of preeclampsia: a systematic review. Am J Obstet Gynecol. 1999;181:1026–1035
- . The paradoxical effect of smoking in preeclamptic pregnancies: smoking reduces the incidence but increases the rates of perinatal mortality, abruptio placentae and intrauterine growth restriction. Am J Obstet Gynecol. 1997;177:156–161
- Maternal smoking and preeclampsia. J Reprod Med. 2000;45:727–732
- . Effects of smoking during pregnancy: five meta-analyses. Am J Prev Med. 1999;16:208–215
- . Urinary cotinine concentration confirms the reduced risk of preeclampsia with tobacco exposure. Am J Obstet Gynecol. 1999;18:1192–1196
- . Cigarette smoking in pregnancy and risk of preeclampsia. J Hum Hypertens. 1994;8:771–775
- . Smoking in pregnancy revisited: findings from a large population-based study. Am J Obstet Gynecol. 2005;192:1856–1863
- . The puzzling association between smoking and hypertension during pregnancy. Am J Obstet Gynecol. 1999;181:1407–1413
- . Moderate smoking during pregnancy is associated with a reduced risk of preeclampsia. Acta Obstet Gynecol Scand. 1999;78:693–697
- . Cigarette smoking and pregnancy II: vascular effects. Placenta. 1999;20:273–279
- . Direct placental effects of cigarette smoke protect women from preeclampsia: the specific roles of carbon monoxide and antioxidant systems in the placenta. Med Hypotheses. 2005;64:17–27
- . Pre-eclampsia. Lancet. 2005;365:785–799
- . Association between maternal smoking and low birth weight in Switzerland: the EDEN study. Swiss Med Wkly. 2005;135:525–530
- . Prenatal smoking and risk of intrapartum stillbirth. Arch Environ Occup Health. 2007;62:87–92
- . Maternal complications and perinatal outcomes associated with gestational hypertension and severe preeclampsia in Taiwanese women. J Formos Med Assoc. 2008;107:129–138
- . Preeclampsia and preterm birth subtypes in Nova Scotia, 1986 to 1992. Am J Perinatol. 1997;14:17–23
- . Clinical presentation and risk factors of placental abruption. Acta Obstet Gynecol Scand. 2006;85:700–705
- . Threshold of metabolic acidosis associated with newborn complications. Am J Obstet Gynecol. 1997;177:1391–1394
- . Preeclampsia as a cause of preterm and late preterm (near-term) births. Semin Perinatol. 2006;30:16–19
- . Trends in fetal and infant survival following preeclampsia. JAMA. 2006;296:1357–1362
- . The incidence of pre-eclamptic toxaemia in smokers and non-smokers. Lancet. 1968;1:994–995
- . A community study of smoking in pregnancy. J Obstet Gynaecol Br Commonw. 1972;79:1057–1073
- . Smoking in moderate/severe preeclampsia worsens pregnancy outcome, but smoking cessation limits the damage. Hypertension. 2008;51:1042–1046
- . Cigarette smoking and preeclampsia: their association and effects on clinical outcomes. J Matern Fetal Med. 2001;10:643–651
- . The effects of smoking and hypertensive disorders on fetal growth. BMC Pregnancy Childbirth. 2006;6:16
- . The effect of smoking and hypertensive disorders on abruptio placentae in Norway 1992-2002. Acta Obstet Gynecol Scand. 2007;86:304–309
- Effectiveness of a pregnancy cessation program. J Obstet Gynecol Neonat Nurs. 1992;21:385–392
- . Facilitating changes in perinatal smoking: the impact of a stage-based workshop for care-providers in British Columbia. Can J Public Health. 2002;93:285–290
- . Pharmacotherapy for smoking cessation during pregnancy. Nicotine Tob Res. 2004;6:S189–S202
- . The epidemiology of smoking: health consequences and benefits of cessation. Drugs. 2002;62(suppl2):1–9
- . Smoking cessation for pregnant women: current Canadian programs and future development. Can J Public Health. 2004;95:278–280
- . Stopping smoking during pregnancy: are we on the right track?. Can J Public Health. 2003;4:297–299
- Smoking before pregnancy and the risk of gestational hypertension and preeclampsia. Am J Obstet Gynecol. 2002;186:1035–1040
Supported by the Ottawa Health Research Institute.
Cite this article as: Miller EC, Cao H, Wen SW, et al. The risk of adverse pregnancy outcomes is increased in preeclamptic women who smoke compared with nonpreeclamptic women who do not smoke. Am J Obstet Gynecol 2010;203:334.e1-8.
PII: S0002-9378(10)00607-1
doi:10.1016/j.ajog.2010.05.020
© 2010 Mosby, Inc. All rights reserved.
Volume 203, Issue 4 , Pages 334.e1-334.e8, October 2010
