- Email: [email protected]
Is the association of maternal smoking and pregnancy-induced hypertension dependent on fetal growth?
Research
www. AJOG.org
OBSTETRICS
Is the association of maternal smoking and pregnancyinduced hypertension dependent on fetal growth? Morgan R. Peltier, PhD; Cande V. Ananth, PhD, MPH OBJECTIVE: The risk of pregnancy-induced hypertension (PIH) is decreased by smoking, but the mechanisms remain unclear. Our objective was to determine whether this association is dependent on decreased fetal growth. METHODS: A population-based, retrospective cohort study in the
United States was performed consisting of nulliparous women who delivered a singleton birth (n ⫽ 8,025,295) between 1995 and 2002. Fetal growth was defined as birthweight for gestational age and characterized as less than 1, 1-2, 3-4, 5-9, 10-19, . . ., 90 or greater centiles. Risk and relative risk of PIH before and after adjusting for confounders were estimated.
RESULTS: Smoking was associated with decreased risk of PIH with up to a 46% decreased risk of PIH for growth-restricted babies (less than the 10th centile). This association, however, decreased with increasing birthweight centile and was nonsignificant at 20th or greater centile among heavy smokers, at 60th or greater centile for moderate, and 80th or greater centile for light smokers. CONCLUSION: Smoking was primarily associated with decreased risk of PIH among growth-restricted babies.
Key Words: fetal growth, pregnancy-induced hypertension, preeclampsia, smoking
Cite this article as: Peltier MR, Ananth CV. Is the association of maternal smoking and pregnancy-induced hypertension dependent on fetal growth? Am J Obstet Gynecol 2007;196;532.e1-532.e6.
P
regnancy-induced hypertension is a common cause of maternal morbidity, affecting 5-8% of all pregnancies,1 and presents as gestational hypertension, From the Divisions of Maternal-Fetal Medicine (Dr Peltier) and Epidemiology and Biostatistics (Dr Ananth), Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Medicine and Dentistry New Jersey–Robert Wood Johnson Medical School, New Brunswick, NJ. Received June 15, 2006; revised October 12, 2006; accepted December 28, 2006. Reprints: Cande V. Ananth, PhD, MPH, Division of Epidemiology and Biostatistics, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Medicine and Dentistry New Jersey–Robert Wood Johnson Medical School, 125 Paterson St, New Brunswick, NJ 08901; [email protected]. This work was supported in part by grants from the National Institutes of Health Loan Repayment Program (to M.R.P.), the University of Medicine and Dentistry of New Jersey Research Foundation (to M.R.P.) and the National Institutes of Health, MD (HD038902; to C.V.A.). 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2006.12.037
532.e1
preeclampsia, or eclampsia. This condition is associated with inhibited fetal growth and is a significant cause of preterm birth. A number of maternal characteristics such as nulliparity, black race, maternal education, and increased body mass index have been reported to increase a woman’s risk for preeclampsia.2 Smoking, however, decreases the risk of preeclampsia and pregnancy-induced hypertension.3-6 The observation that women who quit smoking prior to pregnancy are not at decreased risk for developing preeclampsia or pregnancy-induced hypertension7 suggests that the effects are perhaps transient and pregnancy specific. This association is puzzling because smoking is a consistently reported risk factor for a variety of adverse reproductive8 and pregnancy outcomes.9,10 Although a number of biochemical pathways have been implicated in the pathophysiology of preeclampsia, it is well accepted that preeclampsia is a condition of the placenta.11 This suggests that preeclampsia and perhaps other forms of hypertension are due to the production of 1 or more factors secreted by the placenta that cause endothelial cell dysfunction. It is therefore plausible
American Journal of Obstetrics & Gynecology JUNE 2007
that smoking affects the placenta in such a way to inhibit or prevent the production of these factors. We hypothesized that alterations in placental function caused by smoking may be responsible for the supposedly protective effect of smoking on risk of pregnancyinduced hypertension. Therefore, we evaluated whether there is an association between hypertensive complications of pregnancy and placental function as reflected by birthweight.
M ATERIALS AND M ETHODS The United States’ natality and fetal death data files were utilized for analysis, comprised of all births between 1995 and 2002. These data are compiled and assembled by the National Center for Health Statistics (NCHS) of the U.S. Centers for Disease Control and Prevention. The NCHS routinely links birth and infant death records that are provided by individual states under the Vital Statistics Cooperative Program.12 The linked data include information on maternal characteristics, medical and obstetric history, and complications of pregnancy as well as fetal and infant outcomes.13 Natality and fetal death data are routinely recorded on the live birth and
Obstetrics
www.AJOG.org
Research
TABLE 1
Distribution of maternal characteristics among women with pregnancies complicated by pregnancy-induced hypertension and small for gestational age Non-SGA Number of births
SGA
Normotensive
PIH
Normotensive
PIH
6,708,749
377,643
852,541
86,362
................................................................................................................................................................................................................................................................................................................................................................................
Maternal age (y)*
24.2 ⫾ 5.9
24.2 ⫾ 5.9
23.1 ⫾ 6.0
23.7 ⫾ 6.0
................................................................................................................................................................................................................................................................................................................................................................................ †
Smoking (%)
.......................................................................................................................................................................................................................................................................................................................................................................
Nonsmoker
91.5
92.5
84.8
89.1
Light
6.6
5.8
11.2
8.2
Moderate
1.7
1.5
3.6
2.5
Heavy
0.16
0.15
0.40
0.25
....................................................................................................................................................................................................................................................................................................................................................................... ....................................................................................................................................................................................................................................................................................................................................................................... ....................................................................................................................................................................................................................................................................................................................................................................... ................................................................................................................................................................................................................................................................................................................................................................................
Education (y, %)
.......................................................................................................................................................................................................................................................................................................................................................................
Less than 12
22.3
20.1
30.8
24.6
12
30.3
32.3
33.1
34.5
13-15
20.4
23.0
17.5
20.8
....................................................................................................................................................................................................................................................................................................................................................................... ....................................................................................................................................................................................................................................................................................................................................................................... .......................................................................................................................................................................................................................................................................................................................................................................
16 or more
27.0
24.6
18.6
20.1
Black race (%)
14.5
14.3
24.1
23.11
Unmarried status (%)
40.3
38.6
52.4
47.1
No prenatal care (%)
3.3
2.8
4.7
3.6
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
Gestational age (wks)*
39.1 ⫾ 2.5
38.0 ⫾ 2.9
39.1 ⫾ 2.3
37.9 ⫾ 2.7
3,314 ⫾ 543
3,174 ⫾ 679
2,394 ⫾ 419
2,159 ⫾ 503
................................................................................................................................................................................................................................................................................................................................................................................
Birthweight (g)*
................................................................................................................................................................................................................................................................................................................................................................................
PIH, pregnancy-induced hypertension; SGA, small for gestational age. * Mean ⫾ SD. †
Light, moderate, and heavy smoking defined as 1-10, 11-20, and 21 or more cigarettes smoked per day, respectively.
death certificates, respectively, by attendants at the time of delivery.13 Because preeclampsia is not listed as a separate form of gestational hypertension in this data set, all analyses were restricted to pregnancy-induced hypertension, which included preeclampsia and eclampsia. Gestational age on these data files was largely derived from the last menstrual period for over 95% of pregnancies. When the estimated gestational age on the basis of menstrual dates is contradictory to the reported birthweight, a clinical estimate of gestational age (also contained on the vital statistics data) was instead substituted.14 When the day of the menstrual period was missing (but month and year available), the missing gestational age was statistically imputed in less than 5% of births.14 The replacement of clinically estimated gestational age and the imputation are both performed by the NCHS consistently for all the years examined in this study.13,14
Self-reported maternal smoking during pregnancy was categorized as light (1-10 cigarettes smoked per day), moderate (11-20 cigarettes smoked per day), and heavy (21 or more cigarettes smoked per day). Data on smoking status prior to pregnancy, across trimesters during pregnancy, and quit patterns during pregnancy were not recorded. Fetal growth was assessed in birthweight centiles and was defined as sex-specific birthweight less than 1, 1-2, 3-4, 5-9, 1019, 20-29, . . . 90th or greater centile for gestational age. These cut-offs were derived from all 1995-2002 singleton live births in the United States (ie, internal standards). From a total of 32,428,118 live births and stillbirths in the United States between 1995 and 2002, we sequentially excluded multiple gestations (n ⫽ 1,001,132), missing birthweight or gestational age (n ⫽ 600,703), gestational age less than 22 weeks (n ⫽ 93,700), or ges-
tational age 45 weeks or longer (n ⫽ 298,033). In addition, women diagnosed with chronic hypertension (in the absence of PIH), diabetes, or renal disease were excluded (n ⫽ 1,113,013). Finally, California, Indiana, New York, and South Dakota did not report data on smoking during pregnancy to the NCHS prior to data compilation, so 5,051,500 births from these states were excluded. Pregnancy-induced hypertension is best studied in nulliparous pregnancies because this condition is frequently confounded with reemergence of chronic hypertension or renal diseases because of the stress of pregnancy. Therefore, we excluded multiparous pregnancies to obtain a final analytic data set of 8,025,295 births. We examined the distributions of sociodemographic and behavioral risk factors for PIH, with the analysis stratified based on small for gestational age (sex-specific birthweight less than the
JUNE 2007 American Journal of Obstetrics & Gynecology
532.e2
Research
Obstetrics
www.AJOG.org
FIGURE 1
Association of maternal smoking with pregnancy-induced hypertension
A
B 16
Heavy Smoking
Non-Smoking Light Smoking Moderate Smoking Heavy Smoking
14 12
PIH
Moderate Smoking Light Smoking
10 8 6
NonSmoking
4 2
0.8
0.9
1.0
<1 1- 3- 5- 10-20-30-40-50-60-70-80-90+
RR (95% CI) for PIH
Fetal growth centile
Association of smoking on risk of pregnancy-induced hypertension. A, adjusted relative risk (95% confidence interval). B, Rates of PIH, stratified by fetal growth centile.
10th percentile for gestational age). Relative risk was used to describe the association between maternal smoking (none, light, moderate, or heavy) and PIH. We fitted multivariable logistic regression models to estimate the ad-
justed odds ratio and 95% percent confidence interval for association between smoking and PIH. Because the rate of PIH was fairly low, odds ratio provides a good approximation to the relative risk.
TABLE 2
Association between maternal smoking and pregnancy-induced hypertension within strata of fetal growth centiles: adjusted relative risk with 95% confidence interval Adjusted relative risk (95% confidence interval) for PIH* Fetal growth centile
Light smoker†
Moderate smoker†
Heavy smoker†
Less than 1
0.74 (0.69 to 0.79)
0.61 (0.55 to 0.68)
0.54 (0.40 to 0.73)
1-2
0.66 (0.63 to 0.70)
0.57 (0.52 to 0.63)
0.59 (0.46 to 0.77)
3-4
0.69 (0.65 to 0.73)
0.59 (0.53 to 0.66)
0.62 (0.45 to 0.86)
5-9
0.69 (0.66 to 0.72)
0.70 (0.65 to 0.75)
0.55 (0.43 to 0.71)
10-19
0.75 (0.73 to 0.78)
0.71 (0.67 to 0.75)
0.78 (0.68 to 0.95)
20-29
0.80 (0.77 to 0.84)
0.74 (0.68 to 0.79)
0.79 (0.63 to 1.00)
30-39
0.84 (0.80 to 0.88)
0.83 (0.77 to 0.90)
0.84 (0.65 to 1.09)
40-49
0.86 (0.83 to 0.90)
0.81 (0.74 to 0.88)
0.89 (0.67 to 1.17)
50-59
0.89 (0.85 to 0.93)
0.87 (0.80 to 0.96)
0.83 (0.62 to 1.12)
60-69
0.86 (0.82 to 0.91)
0.96 (0.87 to 1.05)
0.99 (0.74 to 1.32)
70-79
0.93 (0.89 to 0.99)
0.98 (0.88 to 1.08)
1.10 (0.82 to 1.50)
80-89
0.95 (0.90 to 1.00)
1.07 (0.97 to 1.18)
0.87 (0.61 to 1.25)
90 or greater
1.01 (0.96 to 1.06)
1.01 (0.90 to 1.12)
1.19 (0.86 to 1.63)
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
GHTN, gestational hypertension. * Relative risks were adjusted for maternal age, gravida, race/ethnicity, education, marital status, prenatal care, and period of birth. †
Light, moderate, and heavy smoking defined as 1-10, 11-20, and 21 or more cigarettes smoked per day, respectively.
532.e3
American Journal of Obstetrics & Gynecology JUNE 2007
To address issues of confounding, we adjusted the logistic regression models for maternal age (grouped in 5-year intervals as less than 20, 20-24, 25-29, 3034, 35-39, 40-44, and 45 years or more), maternal education (less than 12, 12, 1315, and 16 years or more of completed schooling), maternal race (black, white, and others), marital status (single vs married), and lack of prenatal care. In addition, we adjusted the analyses for birth period (in single years) to minimize any bias because of temporal changes in the incidence of PIH over the study period. Finally, all analyses examining the association between maternal smoking and the risk of PIH were stratified on the basis of fetal growth centiles and birthweight. Statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC). The study was approved by the Institutional Review Board of the University of Medicine and Dentistry New Jersey–Robert Wood Johnson Medical School.
R ESULTS The distributions of maternal age, maternal education, race, marital status, and prenatal care were different between normotensive women and those with PIH on the basis of small-for-gestational-age status (Table 1). Multiple logistic regression analysis suggested that all of these factors and year of birth significantly altered the risk of PIH. Smoking was associated with a decreased risk of PIH, but heavy smokers had a risk of developing PIH that was similar to light smokers (Figure 1, A). The association of smoking with lower rates of PIH differed with fetal growth centile (Figure 1, B). Smoking was associated with a lower risk of PIH for babies born at less than 1 percentile, but this association vanished at 20th or greater, 60th or greater, and 80th or greater centiles for heavy, moderate, and light smokers, respectively (Table 2). Rates and risk of PIH by gestational age across different smoking categories are shown in Figure 2, B. Smoking was associated with a reduced risk for PIH among babies born at term (37-44 weeks), moderately preterm (32-36 weeks), or very preterm (22-31 weeks).
Obstetrics
www.AJOG.org
FIGURE 2
Association between maternal smoking and pregnancy-induced hyertension within strata of gestational age
B
A Non-Smoking Light Smoking Moderate Smoking Heavy Smoking
12
PIH (%)
10 8 6 4 2
Gestational age (weeks)
14
37-44
Non-Smoking Light Smoking Moderate Smoking Heavy Smoking
32-36
22-31
0 22-31
32-36
0.5 RR (95% CI) of PIH
37-44
Gestational age (weeks)
1.0
Effect of maternal smoking on rates and risk of pregnancy-induced hypertension. A, rates and B, relative risk of pregnancy-induced hypertension for babies born at very preterm (22-31 weeks), preterm (32-36 weeks) and term (37-44 weeks) gestation for light (0-10 cigarettes/day), moderate (11-20 cigarettes/day), and heavy (21 or more cigarettes/day).
Rates of PIH by birthweight across smoking categories are shown in Figure 3. When data were analyzed by birthweight, no association of smoking on risk of PIH could be detected at 3000 g or greater for heavy smokers, 3500 g or greater for moderate smokers, and 4000 g or greater for light smokers (Table 3). Reduced risk of PIH by light and heavy smoking was observed in low birthweight babies (less than 2500 g), whereas among moderate smokFIGURE 3
Effect of maternal smoking on pregnancy-induced hypertension, stratified by birthweight
ing, the association was less robust. For babies born between 500 and 1499 g, no association of moderate smoking with PIH was detected.
C OMMENT As with previous studies,3-6 our analysis confirms that maternal smoking is asso-
Research
ciated with reduced fetal growth as well as risk of PIH. No dose-response effects between level of smoking and risk of PIH was observed in our study, however. This may be due to self-reporting errors including those of rounding errors and general misclassification. Because the effect of smoking on the risk of PIH was strongest for the severely growth-restricted babies, our study suggests that the ability for smoking to reduce the risk of PIH may be dependent on decreased fetal growth. For large babies, smoking offered no “protection” against the risk of PIH. Although we limited our analyses to primiparous women in whom PIH is best studied, similar results were observed for multiparous women (data not shown). The strengths of this study are that it is population based, with a very large number of patients and can yield highly robust and persuasive associations. There is also careful adjustment for a variety of confounders through multivariable regression models. Our findings, however, are limited by several factors that are inherent in large population-based studies. First, lack of consistent standards for defining gestational hypertension may have affected
TABLE 3
Association between maternal smoking and pregnancy-induced hypertension within birthweight strata: adjusted relative risk with 95% confidence interval Adjusted relative risk (95% confidence interval) for PIH* Birthweight (g)
Light smoker†
Moderate smoker†
Heavy smoker†
Less than 500
0.73 (0.57 to 0.94)
0.56 (0.34 to 0.90)
0.40 (0.54 to 2.94)
500-
0.86 (0.77 to 0.95)
0.92 (0.77 to 1.09)
0.39 (0.20 to 0.76)
1000-
0.83 (0.77 to 0.90)
0.94 (0.83 to 1.08)
0.56 (0.36 to 0.86)
1500-
0.74 (0.70 to 0.79)
0.70 (0.63 to 0.77)
0.61 (0.45 to 0.82)
2000-
0.62 (0.60 to 0.65)
0.53 (0.49 to 0.57)
0.51 (0.42 to 0.62)
2500-
0.70 (0.69 to 0.72)
0.65 (0.62 to 0.68)
0.63 (0.54 to 0.73)
3000-
0.85 (0.83 to 0.87)
0.84 (0.78 to 0.85)
0.90 (0.79 to 1.02)
3500-
0.93 (0.91 to 0.96)
0.96 (0.91 to 1.02)
0.96 (0.80 to 1.16)
4000-
0.97 (0.91 to 1.03)
0.94 (0.82 to 1.06)
1.36 (0.96 to 1.92)
4500-
1.09 (0.93 to 1.28)
1.39 (1.03 to 1.88)
0.73 (0.18 to 3.03)
5000 or greater
0.87 (0.50 to 1.52)
1.15 (0.35 to 3.78)
Not estimable
.............................................................................................................................................................................................................................................. 20
Non-smoking Light Smoking Moderate Smoking Heavy Smoking
PIH (%)
15
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
10
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
5
.............................................................................................................................................................................................................................................. 0 <500
500-
1000- 1500- 2000- 2500- 3000- 3500- 4000- 4500- 5000+
Birth weight (g)
Effect of light (0-10 cigarettes/day), moderate (11-20 cigarettes/day), and heavy (21 or more cigarettes/day) smoking on rates of pregnancyinduced hypertension is stratified by birthweight.
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
* See Table 2 for list of confounders. †
Light, moderate, and heavy smoking defined as 1-10, 11-20, and 21 or more cigarettes smoked per day, respectively.
JUNE 2007 American Journal of Obstetrics & Gynecology
532.e4
Research
Obstetrics
our results to some extent. There were also no data available with regard to the severity of gestational hypertension, and we could not distinguish PIH from preeclampsia or gestational hypertension. Another limitation of the data is that level of smoking was based on maternal self-reporting and was not confirmed by biochemical measurements. Previous studies have shown that self-reporting is likely to underestimate the number of smokers in a population.15,16 Therefore, the possibility of misclassification of smoking exposure on our findings cannot be overlooked. Such inaccuracies, however, would bias results toward the null. We did observe the expected association between smoking and low birth weight, suggesting that the self-reported data are accurate enough for the objectives of our study. Other limitations of this study include the possibility that smoking behaviors may have changed during the course of the pregnancy. Body mass index and illicit drug use, factors shown to be associated with preeclampsia, were also unavailable in the vital statistics data. Alcohol consumption was severely underreported. It is also unknown whether the growth restriction preceded the development of pregnancy-induced hypertension or followed it. Moreover, because the exact timing of stillbirths is unclear, it is likely that these births were classified as being growth restricted. Finally, there is no information about placental weight or potential pathologies that may have been present in tissues of women who smoked at the growth percentiles studied. Previous studies have shown that smoking can have significant effects on placental development and function.17,18 Many studies have shown that smoking has little or no effect on placental weight19,20 but may increase the ratio of placental to fetal weight, reflecting a decrease in placental function rather than placental growth.21 Placental cells harvested from smokers had decreased amino acid uptake,22 which may cause fetal growth restriction. In addition to transport and uptake of nutrients, endocrine properties of the placenta may also be altered by maternal smoking,23 and it 532.e5
www.AJOG.org is possible that the production of a placental factor(s) that causes pregnancyinduced hypertension may be inhibited by maternal smoking. Although the etiology of gestational hypertension is unknown, there is accumulating evidence that the condition is likely the result of reduced or aberrant angiogenesis.24 Recently a soluble form of the vasoendothelial growth factor receptor (sFlt-1, also known as soluble VEGF receptor 1) that can bind to VEGF as well as placental growth factor has emerged as a likely component of preeclampsia.25 This soluble receptor is present in greater amounts in the placentas and the peripheral blood of preeclamptic women.25,26 Production of sFlt-1 in the placenta is increased by hypoxia,27,28 and increased expression of this gene has been reported in placentas from growth-restricted pregnancies,29 which are at increased risk for pregnancy-induced hypertension. It is possible that smoking damages the placenta to reduce the production of sFlt-1 or perhaps other factors made by the placenta. Recently Powers et al30 reported that sFlt-1 levels were significantly lower in smokers than nonsmokers from 15 to 35 weeks’ gestation, 35 weeks’ gestation until term, and postpartum in normotensive women.30 Furthermore, extracts of cigarette smoke decreased the production of sFlt-1 by placental explants in vitro.31 More studies are needed to determine whether smoking-induced changes in sFlt-1 vary with fetal growth and preeclampsia.30 Although a reduction in placental sFfl-1 production seems to be the simplest explanation for the effects of smoking on risk of pregnancy-induced hypertension, it is also possible that smoking may be affecting other pathways (eg, via maternal vascular endothelium) in parallel. In summary, our study demonstrates that the association between smoking and risk of PIH varies with fetal size as well as gestational age at delivery. Further studies are required, however, to determine whether these effects are specific to PIH or whether they extend to preeclampsia as well. The role of sFlt-1 or other regulators of angiogenesis at the
American Journal of Obstetrics & Gynecology JUNE 2007
maternal-fetal interface on these associf ations needs evaluation.
ACKNOWLEDGMENT The authors thank Anthony Vintzileos, MD, for critically reviewing the manuscript and offering several comments that improved it.
REFERENCES 1. Granger JP, Alexander BT, Bennett WA, Khalil RA. Pathophysiology of pregnancy-induced hypertension. Am J Hypertens 2001;14:178S-85S. 2. Mittendorf R, Lain KY, Williams MA, Walker CK. Preeclampsia. A nested, case-control study of risk factors and their interactions. J Reprod Med 1996;41:491-6. 3. Mortensen JT, Thulstrup AM, Larsen H, Moller M, Sorensen HT. Smoking, sex of the offspring, and risk of placental abruption, placenta previa, and preeclampsia: a populationbased cohort study. Acta Obstet Gynecol Scand 2001;80:894-8. 4. Marcoux S, Brisson J, Fabia J. The effect of cigarette smoking on the risk of preeclampsia and gestational hypertension. Am J Epidemiol 1989;130:950-7. 5. Misra DP, Kiely JL. The effect of smoking on the risk of gestational hypertension. Early Hum Dev 1995;40:95-107. 6. Zhang J, Klebanoff MA, Levine RJ, Puri M, Moyer P. The puzzling association between smoking and hypertension during pregnancy. Am J Obstet Gynecol 1999;181:1407-13. 7. England LJ, Levine RJ, Qian C, et al. Smoking before pregnancy and risk of gestational hypertension and preeclampsia. Am J Obstet Gynecol 2002;186:1035-40. 8. Kmietowicz Z. Smoking is causing impotence, miscarriages, and infertility. BMJ 2004;328:364. 9. Cnattingius S, Granath F, Petersson G, Harlow BL. The influence of gestational age and smoking habits on the risk of subsequent preterm deliveries. N Engl J Med 1999;341:943-8. 10. Cnattingius S, Lambe M. Trends in smoking and overweight during pregnancy: prevalence, risks of pregnancy complications, and adverse pregnancy outcomes. Semin Perinatol 2002;26:286-95. 11. Myatt L. Role of placenta in preeclampsia. Endocrine 2002;19:103-11. 12. National Center for Health Statistics: 19952002 linked birth/infant death data set. NCHS. Hyattsville, MD: United States Department of Health and Human Services, Centers for Disease Control and Prevention. 13. Taffel SM, Ventura SJ, Gay GA. Revised U.S. certificate of birth-new opportunities for research on birth outcome. Birth 1989;16: 188-93.
Obstetrics
www.AJOG.org 14. Taffel S, Johnson D, Heuse R. A method of imputing length of gestation on birth certificates. Vital Health Stat 2 1982;93:1-11. 15. Britton GR, Brinthaupt J, Stehle JM, James GD. Comparison of self-reported smoking and urinary cotinine levels in a rural pregnant population. J Obstet Gynecol Neonatal Nurs 2004;33:306-11. 16. Boyd NR, Windsor RA, Perkins LL, Lowe JB. Quality of measurement of smoking status by self-report and saliva cotinine among pregnant women. Matern Child Health J 1998;2:77-83. 17. Spira A, Philippe E, Spira N, Dreyfus J, Schwartz D. Smoking during pregnancy and placental pathology. Biomedicine 1977;27: 266-70. 18. Naeye RL. Effects of maternal cigarette smoking on the fetus and placenta. Br J Obstet Gynaecol 1978;85:732-7. 19. Spira A, Spira N, Goujard J, Schwartz D. Smoking during pregnancy and placental weight: a multivariate analysis on 3759 cases. J Perinat Med 1975;3:237-41. 20. Christianson RE. Gross differences observed in the placentas of smokers and nonsmokers. Am J Epidemiol 1979;110:178-87.
21. Wingerd J, Christianson R, Lovitt WV, Schoen EJ. Placental ratio in white and black women: relation to smoking and anemia. Am J Obstet Gynecol 1976;124:671-5. 22. Pastrakuljic A, Derewlany LO, Koren G. Maternal cocaine use and cigarette smoking in pregnancy in relation to amino acid transport and fetal growth. Placenta 1999;20:499-512. 23. Mochizuki M, Maruo T, Masuko K, Ohtsu T. Effects of smoking on fetoplacental-maternal system during pregnancy. Am J Obstet Gynecol 1984;149:413-20. 24. Bdolah Y, Sukhatme VP, Karumanchi SA. Angiogenic imbalance in the pathophysiology of preeclampsia: newer insights. Semin Nephrol 2004;24:548-56. 25. Maynard SE, Min JY, Merchan J, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003;111:649-58. 26. Koga K, Osuga Y, Yoshino O, et al. Elevated serum soluble vascular endothelial growth factor receptor 1 (sVEGFR-1) levels in women with preeclampsia. J Clin Endocrinol Metab 2003;88:2348-51.
Research
27. Nagamatsu T, Fujii T, Kusumi M, et al. Cytotrophoblasts up-regulate soluble fms-like tyrosine kinase-1 expression under reduced oxygen: an implication for the placental vascular development and the pathophysiology of preeclampsia. Endocrinology 2004;145:4838-45. 28. Li H, Gu B, Zhang Y, Lewis DF, Wang Y. Hypoxia-induced increase in soluble Flt-1 production correlates with enhanced oxidative stress in trophoblast cells from the human placenta. Placenta 2005;26:210-7. 29. Nevo O, Xu J, Many A, et al. sFlt-1 expression is increased in placentae from IUGR pregnancies. J Soc Gyn Invest 2006;13(Suppl A): Abstract 671. 30. Powers RW, Roberts JM, Cooper KM, et al. Maternal serum soluble fms-like tyrosine kinase 1 concentrations are not increased in early pregnancy and decrease more slowly postpartum in women who develop preeclampsia. Am J Obstet Gynecol 2005;193:185-91. 31. Mehendale R, Hibbard J, Fazleabas A, Leach R. Effect of cigarette smoke extract on in vitro placental sFlt-1 Secretion. J Soc Gyn Invest 2006;10(Suppl A):Abstract 457.
JUNE 2007 American Journal of Obstetrics & Gynecology
532.e6
www. AJOG.org
OBSTETRICS
Is the association of maternal smoking and pregnancyinduced hypertension dependent on fetal growth? Morgan R. Peltier, PhD; Cande V. Ananth, PhD, MPH OBJECTIVE: The risk of pregnancy-induced hypertension (PIH) is decreased by smoking, but the mechanisms remain unclear. Our objective was to determine whether this association is dependent on decreased fetal growth. METHODS: A population-based, retrospective cohort study in the
United States was performed consisting of nulliparous women who delivered a singleton birth (n ⫽ 8,025,295) between 1995 and 2002. Fetal growth was defined as birthweight for gestational age and characterized as less than 1, 1-2, 3-4, 5-9, 10-19, . . ., 90 or greater centiles. Risk and relative risk of PIH before and after adjusting for confounders were estimated.
RESULTS: Smoking was associated with decreased risk of PIH with up to a 46% decreased risk of PIH for growth-restricted babies (less than the 10th centile). This association, however, decreased with increasing birthweight centile and was nonsignificant at 20th or greater centile among heavy smokers, at 60th or greater centile for moderate, and 80th or greater centile for light smokers. CONCLUSION: Smoking was primarily associated with decreased risk of PIH among growth-restricted babies.
Key Words: fetal growth, pregnancy-induced hypertension, preeclampsia, smoking
Cite this article as: Peltier MR, Ananth CV. Is the association of maternal smoking and pregnancy-induced hypertension dependent on fetal growth? Am J Obstet Gynecol 2007;196;532.e1-532.e6.
P
regnancy-induced hypertension is a common cause of maternal morbidity, affecting 5-8% of all pregnancies,1 and presents as gestational hypertension, From the Divisions of Maternal-Fetal Medicine (Dr Peltier) and Epidemiology and Biostatistics (Dr Ananth), Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Medicine and Dentistry New Jersey–Robert Wood Johnson Medical School, New Brunswick, NJ. Received June 15, 2006; revised October 12, 2006; accepted December 28, 2006. Reprints: Cande V. Ananth, PhD, MPH, Division of Epidemiology and Biostatistics, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Medicine and Dentistry New Jersey–Robert Wood Johnson Medical School, 125 Paterson St, New Brunswick, NJ 08901; [email protected]. This work was supported in part by grants from the National Institutes of Health Loan Repayment Program (to M.R.P.), the University of Medicine and Dentistry of New Jersey Research Foundation (to M.R.P.) and the National Institutes of Health, MD (HD038902; to C.V.A.). 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2006.12.037
532.e1
preeclampsia, or eclampsia. This condition is associated with inhibited fetal growth and is a significant cause of preterm birth. A number of maternal characteristics such as nulliparity, black race, maternal education, and increased body mass index have been reported to increase a woman’s risk for preeclampsia.2 Smoking, however, decreases the risk of preeclampsia and pregnancy-induced hypertension.3-6 The observation that women who quit smoking prior to pregnancy are not at decreased risk for developing preeclampsia or pregnancy-induced hypertension7 suggests that the effects are perhaps transient and pregnancy specific. This association is puzzling because smoking is a consistently reported risk factor for a variety of adverse reproductive8 and pregnancy outcomes.9,10 Although a number of biochemical pathways have been implicated in the pathophysiology of preeclampsia, it is well accepted that preeclampsia is a condition of the placenta.11 This suggests that preeclampsia and perhaps other forms of hypertension are due to the production of 1 or more factors secreted by the placenta that cause endothelial cell dysfunction. It is therefore plausible
American Journal of Obstetrics & Gynecology JUNE 2007
that smoking affects the placenta in such a way to inhibit or prevent the production of these factors. We hypothesized that alterations in placental function caused by smoking may be responsible for the supposedly protective effect of smoking on risk of pregnancyinduced hypertension. Therefore, we evaluated whether there is an association between hypertensive complications of pregnancy and placental function as reflected by birthweight.
M ATERIALS AND M ETHODS The United States’ natality and fetal death data files were utilized for analysis, comprised of all births between 1995 and 2002. These data are compiled and assembled by the National Center for Health Statistics (NCHS) of the U.S. Centers for Disease Control and Prevention. The NCHS routinely links birth and infant death records that are provided by individual states under the Vital Statistics Cooperative Program.12 The linked data include information on maternal characteristics, medical and obstetric history, and complications of pregnancy as well as fetal and infant outcomes.13 Natality and fetal death data are routinely recorded on the live birth and
Obstetrics
www.AJOG.org
Research
TABLE 1
Distribution of maternal characteristics among women with pregnancies complicated by pregnancy-induced hypertension and small for gestational age Non-SGA Number of births
SGA
Normotensive
PIH
Normotensive
PIH
6,708,749
377,643
852,541
86,362
................................................................................................................................................................................................................................................................................................................................................................................
Maternal age (y)*
24.2 ⫾ 5.9
24.2 ⫾ 5.9
23.1 ⫾ 6.0
23.7 ⫾ 6.0
................................................................................................................................................................................................................................................................................................................................................................................ †
Smoking (%)
.......................................................................................................................................................................................................................................................................................................................................................................
Nonsmoker
91.5
92.5
84.8
89.1
Light
6.6
5.8
11.2
8.2
Moderate
1.7
1.5
3.6
2.5
Heavy
0.16
0.15
0.40
0.25
....................................................................................................................................................................................................................................................................................................................................................................... ....................................................................................................................................................................................................................................................................................................................................................................... ....................................................................................................................................................................................................................................................................................................................................................................... ................................................................................................................................................................................................................................................................................................................................................................................
Education (y, %)
.......................................................................................................................................................................................................................................................................................................................................................................
Less than 12
22.3
20.1
30.8
24.6
12
30.3
32.3
33.1
34.5
13-15
20.4
23.0
17.5
20.8
....................................................................................................................................................................................................................................................................................................................................................................... ....................................................................................................................................................................................................................................................................................................................................................................... .......................................................................................................................................................................................................................................................................................................................................................................
16 or more
27.0
24.6
18.6
20.1
Black race (%)
14.5
14.3
24.1
23.11
Unmarried status (%)
40.3
38.6
52.4
47.1
No prenatal care (%)
3.3
2.8
4.7
3.6
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
Gestational age (wks)*
39.1 ⫾ 2.5
38.0 ⫾ 2.9
39.1 ⫾ 2.3
37.9 ⫾ 2.7
3,314 ⫾ 543
3,174 ⫾ 679
2,394 ⫾ 419
2,159 ⫾ 503
................................................................................................................................................................................................................................................................................................................................................................................
Birthweight (g)*
................................................................................................................................................................................................................................................................................................................................................................................
PIH, pregnancy-induced hypertension; SGA, small for gestational age. * Mean ⫾ SD. †
Light, moderate, and heavy smoking defined as 1-10, 11-20, and 21 or more cigarettes smoked per day, respectively.
death certificates, respectively, by attendants at the time of delivery.13 Because preeclampsia is not listed as a separate form of gestational hypertension in this data set, all analyses were restricted to pregnancy-induced hypertension, which included preeclampsia and eclampsia. Gestational age on these data files was largely derived from the last menstrual period for over 95% of pregnancies. When the estimated gestational age on the basis of menstrual dates is contradictory to the reported birthweight, a clinical estimate of gestational age (also contained on the vital statistics data) was instead substituted.14 When the day of the menstrual period was missing (but month and year available), the missing gestational age was statistically imputed in less than 5% of births.14 The replacement of clinically estimated gestational age and the imputation are both performed by the NCHS consistently for all the years examined in this study.13,14
Self-reported maternal smoking during pregnancy was categorized as light (1-10 cigarettes smoked per day), moderate (11-20 cigarettes smoked per day), and heavy (21 or more cigarettes smoked per day). Data on smoking status prior to pregnancy, across trimesters during pregnancy, and quit patterns during pregnancy were not recorded. Fetal growth was assessed in birthweight centiles and was defined as sex-specific birthweight less than 1, 1-2, 3-4, 5-9, 1019, 20-29, . . . 90th or greater centile for gestational age. These cut-offs were derived from all 1995-2002 singleton live births in the United States (ie, internal standards). From a total of 32,428,118 live births and stillbirths in the United States between 1995 and 2002, we sequentially excluded multiple gestations (n ⫽ 1,001,132), missing birthweight or gestational age (n ⫽ 600,703), gestational age less than 22 weeks (n ⫽ 93,700), or ges-
tational age 45 weeks or longer (n ⫽ 298,033). In addition, women diagnosed with chronic hypertension (in the absence of PIH), diabetes, or renal disease were excluded (n ⫽ 1,113,013). Finally, California, Indiana, New York, and South Dakota did not report data on smoking during pregnancy to the NCHS prior to data compilation, so 5,051,500 births from these states were excluded. Pregnancy-induced hypertension is best studied in nulliparous pregnancies because this condition is frequently confounded with reemergence of chronic hypertension or renal diseases because of the stress of pregnancy. Therefore, we excluded multiparous pregnancies to obtain a final analytic data set of 8,025,295 births. We examined the distributions of sociodemographic and behavioral risk factors for PIH, with the analysis stratified based on small for gestational age (sex-specific birthweight less than the
JUNE 2007 American Journal of Obstetrics & Gynecology
532.e2
Research
Obstetrics
www.AJOG.org
FIGURE 1
Association of maternal smoking with pregnancy-induced hypertension
A
B 16
Heavy Smoking
Non-Smoking Light Smoking Moderate Smoking Heavy Smoking
14 12
PIH
Moderate Smoking Light Smoking
10 8 6
NonSmoking
4 2
0.8
0.9
1.0
<1 1- 3- 5- 10-20-30-40-50-60-70-80-90+
RR (95% CI) for PIH
Fetal growth centile
Association of smoking on risk of pregnancy-induced hypertension. A, adjusted relative risk (95% confidence interval). B, Rates of PIH, stratified by fetal growth centile.
10th percentile for gestational age). Relative risk was used to describe the association between maternal smoking (none, light, moderate, or heavy) and PIH. We fitted multivariable logistic regression models to estimate the ad-
justed odds ratio and 95% percent confidence interval for association between smoking and PIH. Because the rate of PIH was fairly low, odds ratio provides a good approximation to the relative risk.
TABLE 2
Association between maternal smoking and pregnancy-induced hypertension within strata of fetal growth centiles: adjusted relative risk with 95% confidence interval Adjusted relative risk (95% confidence interval) for PIH* Fetal growth centile
Light smoker†
Moderate smoker†
Heavy smoker†
Less than 1
0.74 (0.69 to 0.79)
0.61 (0.55 to 0.68)
0.54 (0.40 to 0.73)
1-2
0.66 (0.63 to 0.70)
0.57 (0.52 to 0.63)
0.59 (0.46 to 0.77)
3-4
0.69 (0.65 to 0.73)
0.59 (0.53 to 0.66)
0.62 (0.45 to 0.86)
5-9
0.69 (0.66 to 0.72)
0.70 (0.65 to 0.75)
0.55 (0.43 to 0.71)
10-19
0.75 (0.73 to 0.78)
0.71 (0.67 to 0.75)
0.78 (0.68 to 0.95)
20-29
0.80 (0.77 to 0.84)
0.74 (0.68 to 0.79)
0.79 (0.63 to 1.00)
30-39
0.84 (0.80 to 0.88)
0.83 (0.77 to 0.90)
0.84 (0.65 to 1.09)
40-49
0.86 (0.83 to 0.90)
0.81 (0.74 to 0.88)
0.89 (0.67 to 1.17)
50-59
0.89 (0.85 to 0.93)
0.87 (0.80 to 0.96)
0.83 (0.62 to 1.12)
60-69
0.86 (0.82 to 0.91)
0.96 (0.87 to 1.05)
0.99 (0.74 to 1.32)
70-79
0.93 (0.89 to 0.99)
0.98 (0.88 to 1.08)
1.10 (0.82 to 1.50)
80-89
0.95 (0.90 to 1.00)
1.07 (0.97 to 1.18)
0.87 (0.61 to 1.25)
90 or greater
1.01 (0.96 to 1.06)
1.01 (0.90 to 1.12)
1.19 (0.86 to 1.63)
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
GHTN, gestational hypertension. * Relative risks were adjusted for maternal age, gravida, race/ethnicity, education, marital status, prenatal care, and period of birth. †
Light, moderate, and heavy smoking defined as 1-10, 11-20, and 21 or more cigarettes smoked per day, respectively.
532.e3
American Journal of Obstetrics & Gynecology JUNE 2007
To address issues of confounding, we adjusted the logistic regression models for maternal age (grouped in 5-year intervals as less than 20, 20-24, 25-29, 3034, 35-39, 40-44, and 45 years or more), maternal education (less than 12, 12, 1315, and 16 years or more of completed schooling), maternal race (black, white, and others), marital status (single vs married), and lack of prenatal care. In addition, we adjusted the analyses for birth period (in single years) to minimize any bias because of temporal changes in the incidence of PIH over the study period. Finally, all analyses examining the association between maternal smoking and the risk of PIH were stratified on the basis of fetal growth centiles and birthweight. Statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC). The study was approved by the Institutional Review Board of the University of Medicine and Dentistry New Jersey–Robert Wood Johnson Medical School.
R ESULTS The distributions of maternal age, maternal education, race, marital status, and prenatal care were different between normotensive women and those with PIH on the basis of small-for-gestational-age status (Table 1). Multiple logistic regression analysis suggested that all of these factors and year of birth significantly altered the risk of PIH. Smoking was associated with a decreased risk of PIH, but heavy smokers had a risk of developing PIH that was similar to light smokers (Figure 1, A). The association of smoking with lower rates of PIH differed with fetal growth centile (Figure 1, B). Smoking was associated with a lower risk of PIH for babies born at less than 1 percentile, but this association vanished at 20th or greater, 60th or greater, and 80th or greater centiles for heavy, moderate, and light smokers, respectively (Table 2). Rates and risk of PIH by gestational age across different smoking categories are shown in Figure 2, B. Smoking was associated with a reduced risk for PIH among babies born at term (37-44 weeks), moderately preterm (32-36 weeks), or very preterm (22-31 weeks).
Obstetrics
www.AJOG.org
FIGURE 2
Association between maternal smoking and pregnancy-induced hyertension within strata of gestational age
B
A Non-Smoking Light Smoking Moderate Smoking Heavy Smoking
12
PIH (%)
10 8 6 4 2
Gestational age (weeks)
14
37-44
Non-Smoking Light Smoking Moderate Smoking Heavy Smoking
32-36
22-31
0 22-31
32-36
0.5 RR (95% CI) of PIH
37-44
Gestational age (weeks)
1.0
Effect of maternal smoking on rates and risk of pregnancy-induced hypertension. A, rates and B, relative risk of pregnancy-induced hypertension for babies born at very preterm (22-31 weeks), preterm (32-36 weeks) and term (37-44 weeks) gestation for light (0-10 cigarettes/day), moderate (11-20 cigarettes/day), and heavy (21 or more cigarettes/day).
Rates of PIH by birthweight across smoking categories are shown in Figure 3. When data were analyzed by birthweight, no association of smoking on risk of PIH could be detected at 3000 g or greater for heavy smokers, 3500 g or greater for moderate smokers, and 4000 g or greater for light smokers (Table 3). Reduced risk of PIH by light and heavy smoking was observed in low birthweight babies (less than 2500 g), whereas among moderate smokFIGURE 3
Effect of maternal smoking on pregnancy-induced hypertension, stratified by birthweight
ing, the association was less robust. For babies born between 500 and 1499 g, no association of moderate smoking with PIH was detected.
C OMMENT As with previous studies,3-6 our analysis confirms that maternal smoking is asso-
Research
ciated with reduced fetal growth as well as risk of PIH. No dose-response effects between level of smoking and risk of PIH was observed in our study, however. This may be due to self-reporting errors including those of rounding errors and general misclassification. Because the effect of smoking on the risk of PIH was strongest for the severely growth-restricted babies, our study suggests that the ability for smoking to reduce the risk of PIH may be dependent on decreased fetal growth. For large babies, smoking offered no “protection” against the risk of PIH. Although we limited our analyses to primiparous women in whom PIH is best studied, similar results were observed for multiparous women (data not shown). The strengths of this study are that it is population based, with a very large number of patients and can yield highly robust and persuasive associations. There is also careful adjustment for a variety of confounders through multivariable regression models. Our findings, however, are limited by several factors that are inherent in large population-based studies. First, lack of consistent standards for defining gestational hypertension may have affected
TABLE 3
Association between maternal smoking and pregnancy-induced hypertension within birthweight strata: adjusted relative risk with 95% confidence interval Adjusted relative risk (95% confidence interval) for PIH* Birthweight (g)
Light smoker†
Moderate smoker†
Heavy smoker†
Less than 500
0.73 (0.57 to 0.94)
0.56 (0.34 to 0.90)
0.40 (0.54 to 2.94)
500-
0.86 (0.77 to 0.95)
0.92 (0.77 to 1.09)
0.39 (0.20 to 0.76)
1000-
0.83 (0.77 to 0.90)
0.94 (0.83 to 1.08)
0.56 (0.36 to 0.86)
1500-
0.74 (0.70 to 0.79)
0.70 (0.63 to 0.77)
0.61 (0.45 to 0.82)
2000-
0.62 (0.60 to 0.65)
0.53 (0.49 to 0.57)
0.51 (0.42 to 0.62)
2500-
0.70 (0.69 to 0.72)
0.65 (0.62 to 0.68)
0.63 (0.54 to 0.73)
3000-
0.85 (0.83 to 0.87)
0.84 (0.78 to 0.85)
0.90 (0.79 to 1.02)
3500-
0.93 (0.91 to 0.96)
0.96 (0.91 to 1.02)
0.96 (0.80 to 1.16)
4000-
0.97 (0.91 to 1.03)
0.94 (0.82 to 1.06)
1.36 (0.96 to 1.92)
4500-
1.09 (0.93 to 1.28)
1.39 (1.03 to 1.88)
0.73 (0.18 to 3.03)
5000 or greater
0.87 (0.50 to 1.52)
1.15 (0.35 to 3.78)
Not estimable
.............................................................................................................................................................................................................................................. 20
Non-smoking Light Smoking Moderate Smoking Heavy Smoking
PIH (%)
15
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
10
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
5
.............................................................................................................................................................................................................................................. 0 <500
500-
1000- 1500- 2000- 2500- 3000- 3500- 4000- 4500- 5000+
Birth weight (g)
Effect of light (0-10 cigarettes/day), moderate (11-20 cigarettes/day), and heavy (21 or more cigarettes/day) smoking on rates of pregnancyinduced hypertension is stratified by birthweight.
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
* See Table 2 for list of confounders. †
Light, moderate, and heavy smoking defined as 1-10, 11-20, and 21 or more cigarettes smoked per day, respectively.
JUNE 2007 American Journal of Obstetrics & Gynecology
532.e4
Research
Obstetrics
our results to some extent. There were also no data available with regard to the severity of gestational hypertension, and we could not distinguish PIH from preeclampsia or gestational hypertension. Another limitation of the data is that level of smoking was based on maternal self-reporting and was not confirmed by biochemical measurements. Previous studies have shown that self-reporting is likely to underestimate the number of smokers in a population.15,16 Therefore, the possibility of misclassification of smoking exposure on our findings cannot be overlooked. Such inaccuracies, however, would bias results toward the null. We did observe the expected association between smoking and low birth weight, suggesting that the self-reported data are accurate enough for the objectives of our study. Other limitations of this study include the possibility that smoking behaviors may have changed during the course of the pregnancy. Body mass index and illicit drug use, factors shown to be associated with preeclampsia, were also unavailable in the vital statistics data. Alcohol consumption was severely underreported. It is also unknown whether the growth restriction preceded the development of pregnancy-induced hypertension or followed it. Moreover, because the exact timing of stillbirths is unclear, it is likely that these births were classified as being growth restricted. Finally, there is no information about placental weight or potential pathologies that may have been present in tissues of women who smoked at the growth percentiles studied. Previous studies have shown that smoking can have significant effects on placental development and function.17,18 Many studies have shown that smoking has little or no effect on placental weight19,20 but may increase the ratio of placental to fetal weight, reflecting a decrease in placental function rather than placental growth.21 Placental cells harvested from smokers had decreased amino acid uptake,22 which may cause fetal growth restriction. In addition to transport and uptake of nutrients, endocrine properties of the placenta may also be altered by maternal smoking,23 and it 532.e5
www.AJOG.org is possible that the production of a placental factor(s) that causes pregnancyinduced hypertension may be inhibited by maternal smoking. Although the etiology of gestational hypertension is unknown, there is accumulating evidence that the condition is likely the result of reduced or aberrant angiogenesis.24 Recently a soluble form of the vasoendothelial growth factor receptor (sFlt-1, also known as soluble VEGF receptor 1) that can bind to VEGF as well as placental growth factor has emerged as a likely component of preeclampsia.25 This soluble receptor is present in greater amounts in the placentas and the peripheral blood of preeclamptic women.25,26 Production of sFlt-1 in the placenta is increased by hypoxia,27,28 and increased expression of this gene has been reported in placentas from growth-restricted pregnancies,29 which are at increased risk for pregnancy-induced hypertension. It is possible that smoking damages the placenta to reduce the production of sFlt-1 or perhaps other factors made by the placenta. Recently Powers et al30 reported that sFlt-1 levels were significantly lower in smokers than nonsmokers from 15 to 35 weeks’ gestation, 35 weeks’ gestation until term, and postpartum in normotensive women.30 Furthermore, extracts of cigarette smoke decreased the production of sFlt-1 by placental explants in vitro.31 More studies are needed to determine whether smoking-induced changes in sFlt-1 vary with fetal growth and preeclampsia.30 Although a reduction in placental sFfl-1 production seems to be the simplest explanation for the effects of smoking on risk of pregnancy-induced hypertension, it is also possible that smoking may be affecting other pathways (eg, via maternal vascular endothelium) in parallel. In summary, our study demonstrates that the association between smoking and risk of PIH varies with fetal size as well as gestational age at delivery. Further studies are required, however, to determine whether these effects are specific to PIH or whether they extend to preeclampsia as well. The role of sFlt-1 or other regulators of angiogenesis at the
American Journal of Obstetrics & Gynecology JUNE 2007
maternal-fetal interface on these associf ations needs evaluation.
ACKNOWLEDGMENT The authors thank Anthony Vintzileos, MD, for critically reviewing the manuscript and offering several comments that improved it.
REFERENCES 1. Granger JP, Alexander BT, Bennett WA, Khalil RA. Pathophysiology of pregnancy-induced hypertension. Am J Hypertens 2001;14:178S-85S. 2. Mittendorf R, Lain KY, Williams MA, Walker CK. Preeclampsia. A nested, case-control study of risk factors and their interactions. J Reprod Med 1996;41:491-6. 3. Mortensen JT, Thulstrup AM, Larsen H, Moller M, Sorensen HT. Smoking, sex of the offspring, and risk of placental abruption, placenta previa, and preeclampsia: a populationbased cohort study. Acta Obstet Gynecol Scand 2001;80:894-8. 4. Marcoux S, Brisson J, Fabia J. The effect of cigarette smoking on the risk of preeclampsia and gestational hypertension. Am J Epidemiol 1989;130:950-7. 5. Misra DP, Kiely JL. The effect of smoking on the risk of gestational hypertension. Early Hum Dev 1995;40:95-107. 6. Zhang J, Klebanoff MA, Levine RJ, Puri M, Moyer P. The puzzling association between smoking and hypertension during pregnancy. Am J Obstet Gynecol 1999;181:1407-13. 7. England LJ, Levine RJ, Qian C, et al. Smoking before pregnancy and risk of gestational hypertension and preeclampsia. Am J Obstet Gynecol 2002;186:1035-40. 8. Kmietowicz Z. Smoking is causing impotence, miscarriages, and infertility. BMJ 2004;328:364. 9. Cnattingius S, Granath F, Petersson G, Harlow BL. The influence of gestational age and smoking habits on the risk of subsequent preterm deliveries. N Engl J Med 1999;341:943-8. 10. Cnattingius S, Lambe M. Trends in smoking and overweight during pregnancy: prevalence, risks of pregnancy complications, and adverse pregnancy outcomes. Semin Perinatol 2002;26:286-95. 11. Myatt L. Role of placenta in preeclampsia. Endocrine 2002;19:103-11. 12. National Center for Health Statistics: 19952002 linked birth/infant death data set. NCHS. Hyattsville, MD: United States Department of Health and Human Services, Centers for Disease Control and Prevention. 13. Taffel SM, Ventura SJ, Gay GA. Revised U.S. certificate of birth-new opportunities for research on birth outcome. Birth 1989;16: 188-93.
Obstetrics
www.AJOG.org 14. Taffel S, Johnson D, Heuse R. A method of imputing length of gestation on birth certificates. Vital Health Stat 2 1982;93:1-11. 15. Britton GR, Brinthaupt J, Stehle JM, James GD. Comparison of self-reported smoking and urinary cotinine levels in a rural pregnant population. J Obstet Gynecol Neonatal Nurs 2004;33:306-11. 16. Boyd NR, Windsor RA, Perkins LL, Lowe JB. Quality of measurement of smoking status by self-report and saliva cotinine among pregnant women. Matern Child Health J 1998;2:77-83. 17. Spira A, Philippe E, Spira N, Dreyfus J, Schwartz D. Smoking during pregnancy and placental pathology. Biomedicine 1977;27: 266-70. 18. Naeye RL. Effects of maternal cigarette smoking on the fetus and placenta. Br J Obstet Gynaecol 1978;85:732-7. 19. Spira A, Spira N, Goujard J, Schwartz D. Smoking during pregnancy and placental weight: a multivariate analysis on 3759 cases. J Perinat Med 1975;3:237-41. 20. Christianson RE. Gross differences observed in the placentas of smokers and nonsmokers. Am J Epidemiol 1979;110:178-87.
21. Wingerd J, Christianson R, Lovitt WV, Schoen EJ. Placental ratio in white and black women: relation to smoking and anemia. Am J Obstet Gynecol 1976;124:671-5. 22. Pastrakuljic A, Derewlany LO, Koren G. Maternal cocaine use and cigarette smoking in pregnancy in relation to amino acid transport and fetal growth. Placenta 1999;20:499-512. 23. Mochizuki M, Maruo T, Masuko K, Ohtsu T. Effects of smoking on fetoplacental-maternal system during pregnancy. Am J Obstet Gynecol 1984;149:413-20. 24. Bdolah Y, Sukhatme VP, Karumanchi SA. Angiogenic imbalance in the pathophysiology of preeclampsia: newer insights. Semin Nephrol 2004;24:548-56. 25. Maynard SE, Min JY, Merchan J, et al. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 2003;111:649-58. 26. Koga K, Osuga Y, Yoshino O, et al. Elevated serum soluble vascular endothelial growth factor receptor 1 (sVEGFR-1) levels in women with preeclampsia. J Clin Endocrinol Metab 2003;88:2348-51.
Research
27. Nagamatsu T, Fujii T, Kusumi M, et al. Cytotrophoblasts up-regulate soluble fms-like tyrosine kinase-1 expression under reduced oxygen: an implication for the placental vascular development and the pathophysiology of preeclampsia. Endocrinology 2004;145:4838-45. 28. Li H, Gu B, Zhang Y, Lewis DF, Wang Y. Hypoxia-induced increase in soluble Flt-1 production correlates with enhanced oxidative stress in trophoblast cells from the human placenta. Placenta 2005;26:210-7. 29. Nevo O, Xu J, Many A, et al. sFlt-1 expression is increased in placentae from IUGR pregnancies. J Soc Gyn Invest 2006;13(Suppl A): Abstract 671. 30. Powers RW, Roberts JM, Cooper KM, et al. Maternal serum soluble fms-like tyrosine kinase 1 concentrations are not increased in early pregnancy and decrease more slowly postpartum in women who develop preeclampsia. Am J Obstet Gynecol 2005;193:185-91. 31. Mehendale R, Hibbard J, Fazleabas A, Leach R. Effect of cigarette smoke extract on in vitro placental sFlt-1 Secretion. J Soc Gyn Invest 2006;10(Suppl A):Abstract 457.
JUNE 2007 American Journal of Obstetrics & Gynecology
532.e6