SEARCH

SEARCH BY CITATION

Keywords:

  • Endothelial function;
  • flow-mediated dilatation;
  • growth restriction;
  • pre-eclampsia;
  • pregnancy;
  • smoking

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

This study examined the relationship of cigarette smoking and endothelial function in pregnant women by comparing smokers with nonsmokers. Endothelial function was assessed at 28–32 weeks of gestation by flow-mediated dilatation (FMD) using ultrasound of the brachial artery. The initial FMD was significantly different between the smoking group (n = 21) at 4.0 ± 2.3, indicating endothelial dysfunction, and the nonsmoking group (n = 20) at 9.7 ± 4.0 (P < 0.001). After smoking, this difference in the groups persisted. Babies who were growth restricted (<10th percentile) had mothers with a significantly lower FMD, that is endothelial dysfunction. This work demonstrates persistent endothelial dysfunction in smoking pregnant women.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

Smoking in pregnancy is known to increase the risk of vascular complications in otherwise low-risk women. In early pregnancy, smoking increases the incidence of miscarriage. These early miscarriages may occur because of apoptosis and decreased invasion of the cytotrophoblast with a resulting increase in placental dysfunction. Once a pregnancy is continuing, smoking increases the risk of preterm birth, placental abruption and placenta praevia. Women who smoke in pregnancy are also at increased risk of delivering babies with intrauterine growth restriction (IUGR) and increased perinatal morbidity and mortality.1 All these complications are thought to result from a reduction in blood flow and decreased perfusion to the placenta with resultant placental injury. Indeed, epidemiological research suggests that smoking in pregnancy induces vascular dysfunction in the uterus.1

Reduced placental perfusion is also thought to be a contributing factor in pre-eclampsia. However, a reduction in placental perfusion alone does not automatically result in pre-eclampsia as failure of spiral artery remodelling and reduced placental perfusion also occurs in IUGR.2 Interestingly, both pre-eclampsia and IUGR are thought to be associated with endothelial dysfunction. The pathogenesis of pre-eclampsia is contentious. One theory is that endothelial dysfunction results from reduced placental perfusion. The addition of a pathological maternal syndrome2 then leads to pre-eclampsia. Another hypothesis suggests that endothelial dysfunction is present pre-pregnancy. Pre-eclampsia then develops after interaction with maternal cytokines, whereas there is no interaction with the maternal metabolic syndrome in IUGR.2 Research has also shown that women who develop pre-eclampsia or deliver an IUGR baby are at increased risk of cardiovascular disease later in life.2

Paradoxically, while pre-eclampsia and IUGR are thought to have a common primary pathophysiology, smoking in pregnancy (which is also associated with IUGR and reduced placental perfusion) reduces the risk of pre-eclampsia by 32%.3 Indeed, the risk of pre-eclampsia decreases as the number of cigarettes smoked daily increases.3 However, smoking in nonpregnant individuals results in reduced dilatation of the vascular endothelium, that is endothelial dysfunction.4 A primary function of the vascular endothelium is to produce nitric oxide (NO), which promotes vasodilatation, whereas endothelial dysfunction occurs when there is a decrease in NO production and decreased vasodilatation.4 We hypothesised that smoking in pregnancy may provide a protective effect on the vascular endothelium by demonstrating no difference in vasodilatory endothelial function when compared with nonsmoking pregnant women. Therefore, the aim of this work was to compare the effects of smoking on endothelial function in pregnant women with nonsmoking pregnant women.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

This study recruited pregnant women who smoked ten or more cigarettes a day and compared them with pregnant women who did not smoke. The women were studied between 28 and 32 weeks of gestation. All women were asked to have a low-fat breakfast and refrain from caffeine drinks. Smoking pregnant women were asked to abstain from cigarettes from midnight. Women with any existing medical conditions were not recruited.

Endothelial function was assessed by the ultrasound technique of flow-mediated dilatation (FMD) as previously described in our normal range study.5 The women had a third-trimester fetal wellbeing ultrasound immediately before the study and had their FMD test commenced at 9 a.m. After 10 minutes of rest, a blood pressure (BP) cuff was placed on the lower arm and inflated for 5 minutes at 200 mmHg to induce reactive hyperaemia. Baseline and post-occlusion brachial artery diameters were measured using a Philips Medical System HDI 5000 ultrasound machine (Philips Ultrasound, Bothell, WA, USA). FMD was calculated using the formula FMD% = ([Post-occlusion artery diameter − baseline artery diameter]/baseline artery diameter) × 100.4 At the completion of the test, the women had a 10- to 15-minute break during which time the smokers could choose to have none, one or two cigarettes. All the participants could have a noncaffeine drink. The FMD test was then repeated in the same manner on all volunteers after 10 minutes of rest to allow for acclimatisation. The break and the repeat test for the nonsmoking women were to measure test-retest reliability. All studies and measurements were performed by a single person (A.E.Q.). Approval for this project was given by both the Area Health Service and University Ethics Committees, and written informed consent was obtained for enrolment.

Statistics were performed using the statistical package SPSS 15.0 (SPSS Inc., Chicago, IL, USA). For normally distributed data, independent t tests were used with data reported as mean ± SD. For analysis of the test-retest in the two groups, a mixed between-within subjects analysis of variance was performed. Categorical data were analysed using chi-square or a Fisher’s exact test if cell size was <5. A significance value of P = 0.05 was used.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

Twenty-two women who smoked during pregnancy and 20 nonsmoking pregnant women were enrolled. One woman who smoked was excluded from analysis as she chose not to have a cigarette in the break. All participants stated no alcohol intake during pregnancy, and only one woman (a smoker) used marijuana occasionally. Women confirmed their consumption of a caffeine-free, low-fat breakfast. The smoking group also confirmed their smoking abstention, with several women commenting that they had remained awake to have a cigarette and coffee at midnight. There was no financial or other incentive for the women to participate in this study.

There were no significant differences in participant characteristics, including physiological parameters, at enrolment between the women who smoked (n = 21) and those who did not smoke (n = 20) except for parity (Table 1). There were less nulliparous women in the smoking group. In all, six women who smoked had complications in a previous pregnancy (Table 1). There were no significant differences in the fetal parameters of head circumference, abdominal circumference and umbilical artery Doppler waveforms at the time of study (Table 1). The mean number of cigarettes smoked per day was 15.7 ± 4.9 (range 10–30), with 19 of the women choosing to smoke one cigarette in the break and 2 choosing to smoke two cigarettes.

Table 1.  Baseline participant characteristics
 Smoker (n = 21)Nonsmoker (n = 20)P value
  • NS, not significant.

  • Values are reported as mean ± SD.

  • *

    Two fetuses IUGR plus pre-eclampsia.

  • **

    One fetus FDIU plus IUGR plus pre-eclampsia.

Maternal age (years)26.6 ± 5.827.5 ± 5.1NS
Nulliparous7/21 (33%)15/20 (75%)0.007
Obstetric history
Pre-eclampsia3/14 (21%)0NS
IUGR*4/14 (29%)0NS
Fetal death in utero (FDIU)**2/14 (14%)0NS
Body mass index28.9 ± 5.929.8 ± 5.6NS
Baseline brachial artery diameter (mm)3.5 ± 0.43.4 ± 0.4NS
Maternal heart rate (bpm)83 ± 9.279 ± 9.3NS
Systolic BP (mmHg)113 ± 11.0111 ± 9.9NS
Diastolic BP (mmHg)70 ± 9.571 ± 8.3NS
Gestational age (weeks)29.3 ± 1.329.1 ± 1.0NS
Umbilical artery Doppler pulsatility index1.1 ± 0.21.1 ± 0.1NS
Fetal head circumference (mm)276 ± 26273 ± 11NS
Fetal abdominal circumference (mm)250 ± 14257 ± 15NS

In the women who smoked, the baseline study FMD was significantly lower compared with those who did not smoke (P < 0.001) (Table 2). At the retest, after the women in the smoking group had a cigarette/s, this significant difference in FMD persisted (P < 0.001) (Table 2). The initial FMD calculated when the women abstained from cigarettes for 9 hours displayed no significant difference compared with the FMD after smoking one to two cigarettes (Table 2). There was also no significant difference in FMD in the nonsmokers between the baseline study and after a 15-minute break (Table 2). Smoking a cigarette significantly increased the physiological variables of heart rate (83 ± 9.3 versus 94 ± 10.7; P < 0.001) and diastolic BP (70 ± 9.5 versus 73 ± 11.6; P = 0.006).

Table 2.  FMD (mean ± SD) in pregnant women who smoke versus pregnant women who are nonsmokers
FMD %Smokers (n = 21)Nonsmokers (n = 20)P value
  • NS, not significant.

  • *

    Between-groups main effect.

  • **

    Within-group interaction effect.

Test4.0 ± 2.39.7 ± 4.0<0.001*
Retest4.4 ± 2.59.6 ± 3.8<0.001*
P valueNS**NS** 

At delivery, there was no significant difference in gestational age between the two groups (smokers: 39.2 ± 1.6 versus nonsmokers: 39.9 ± 1.2 weeks). However, birthweight was significantly decreased in the smoking group (3090 ± 596 g) compared with the nonsmoking group (3501 ± 396 g) (P = 0.014). Growth restriction was evident in the babies of smokers with 38% (8/21) less than the 10th percentile (P = 0.003; Fisher’s exact test) and 24% (5/21) less than the 5th percentile (P = 0.048; Fisher’s exact test). Of particular interest, babies whose birthweight was less than the 10th percentile were born to mothers with a significantly lower FMD of 4.7 ± 2.2 compared with those whose birthweight was greater than the 10th percentile (maternal FMD 7.3 ± 4.6 [P = 0.03]). There were no small-for-gestational-age babies in the nonsmoking group. Only one woman (a smoker) developed gestational hypertension at term.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

This study clearly demonstrates that smoking in pregnancy results in significantly reduced FMD, indicating endothelial dysfunction. We believe that this is the first study to use the technique of FMD to report the effects of smoking on endothelial function in pregnant women. FMD in the nonsmoking women was the same as in our previously reported normal range study.5

There have been a small number of studies on nonpregnant individuals that have demonstrated differing results. One study assessing the effect of smoking on endothelial function demonstrated significantly reduced FMD, indicating endothelial dysfunction with similar FMD results to our work (FMD = 4.0 ± 3.9).4 Two studies have assessed the chronic and acute effect of cigarette smoking by using the test-retest method on nonpregnant participants. The first study demonstrated no difference in FMD between smokers and nonsmokers (both genders) at baseline after an 8-hour break from smoking. A reduced FMD was recorded after both smokers and nonsmokers smoked a cigarette. This study used a mixed study group and had small numbers (n = 27), which were divided into four groups for analysis.6 The second study also had small numbers to assess the effects of smoking a cigarette or using nicotine nasal spray on healthy smoking men (n = 8) and women (n = 8). This study demonstrated no effect from the chronic use of cigarettes on baseline FMD, with the FMD from smokers almost twice that of other studies (FMD = 10.2 ± 4.4 [nasal spray group] and 9.4 ± 3.8 [cigarette group]). A significant reduction in FMD after both treatments was reported.7 However, in this study, the BP cuff was placed on the upper arm,7 and this has been shown to reflect ischaemia, not solely endothelial function.

The endothelium has many functions including control of vascular homeostasis by modulating vascular tone, vessel size and the regulation of blood flow. The endothelium also inhibits inflammation, counteracts the activation of platelets and inhibits proliferation of the vascular smooth muscle.2 This work has concentrated on endothelial function in relation to regulation of vascular tone. Endothelial dysfunction occurs when there is a loss of the functional integrity of the endothelium with a reduction in NO production. When this occurs, there is reduced vasodilation, an increase in the vascular wall inflammatory response and an increase in thrombotic events. Endothelial dysfunction can result from a reduced availability of NO because of altered endothelial nitric oxide synthase expression or accelerated consumption of NO by reactive oxygen species or the reactive nitrogen oxide species.8 FMD is a measure of vascular endothelial function, with reduced FMD indicating endothelial dysfunction.4 It has been postulated that pre-eclampsia results in endothelial dysfunction with resulting vasoconstriction and decreased organ flow from reduced perfusion.2 Of interest, women who deliver an IUGR baby also have an increased risk for coronary artery disease in later life, indicating that these women may also have endothelial dysfunction.2 Our results demonstrating endothelial dysfunction in women with IUGR babies supports this hypothesis.

A small number of laboratory studies have assessed the effects of smoking in pregnancy on the endothelium. Smoking was found to be harmful to the vascular endothelium in a pathological study looking at human umbilical arteries9 and caused degeneration of villous capillary endothelial cells in human placentas.10 These studies and our own study demonstrate that smoking in pregnancy has a deleterious effect on the vascular endothelium.

Cigarette smoke contains many compounds that damage the endothelium.4 Acute nicotine infusion has been shown to cause vasoconstriction and endothelial dysfunction in hamster cheek arterioles.11 Nicotine increases heart rate and BP in humans, with heart rate increasing maximally after the first cigarette of the day. Heart rate then plateaus as tolerance develops.12 Our study demonstrated this increase in maternal heart rate and diastolic BP after the women smoked their first cigarette of the day.

As pre-eclampsia has been associated with endothelial dysfunction and smoking results in endothelial dysfunction, it would be important in future studies to separate women with pre-eclampsia into smoking and nonsmoking groups for analysis. This work also suggests that endothelial dysfunction in the form of reduced vasodilatation is not on its own enough to result in pre-eclampsia as none of the women developed pre-eclampsia. It also raises the question that if pregnant women who smoke have a decreased incidence of pre-eclampsia, is pre-eclampsia solely an endothelial cell disorder? Nearly 40% of the babies of women who smoked in this study were growth restricted. While this number may seem excessive, we did limit the study group to heavy smokers, which could increase the incidence of IUGR. Of particular importance is our finding of a significant relationship between endothelial dysfunction and IUGR in the women who smoked. This supports the hypothesis that IUGR is related to endothelial dysfunction.2 Further studies are planned to determine if endothelial dysfunction persists in pregnant women who have ceased smoking prior to pregnancy or who have reduced their level of smoking to less than five cigarettes a day during their pregnancy.

To obtain ethical approval for this work, limitations were placed on the study design. The women could not be recruited until 28 weeks of gestation because it was considered necessary to allow them time to quit or reduce their smoking. Recruitment was also limited to heavy smokers, defined as ten or more cigarettes a day to ensure we were not asking the women to smoke more than they normally would. Women were also given the choice about whether to smoke or not at the break.

In conclusion, this work has three significant findings. First, it demonstrates that pregnant women who smoke have endothelial dysfunction. Second, using this test-retest method, this endothelial dysfunction was shown to be persistent. Specifically, FMD in the women who smoked was the same after a 9-hour abstention from smoking as it was after smoking one or two cigarettes. Third, this work provides evidence of the relationship between endothelial dysfunction and IUGR. Finally, this study does not support the concept of smoking in pregnancy providing a protective effect on the vascular endothelium. We still have no explanation as to why smoking in pregnancy is associated with a reduced risk of pre-eclampsia. However, the aetiological pathway between endothelial dysfunction, pre-eclampsia and IUGR demands further investigation.

Contribution to authorship

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

A.E.Q. contributed to conception and design of the project, recruited the subjects, collected, collated, analysed and interpreted the data and drafted the article. C.-M.C. contributed to conception and design of the project, analysed and interpreted the data and critically reviewed the article. M.J.P. contributed to conception and design of the project, critically reviewed the article and gave final approval of the version to be published.

Details of ethics approval

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

This study received ethics approval from the Sydney West Area Health Service Ethics Committee (formally known as the Wentworth Area Health Service Ethics Committee), project No. 04/004, approval date 1 March 2004. Ethics approval was also obtained from the Executive Committee of the Human Research Ethics Committee of The University of Sydney, reference number 7635 and approval date 17 August 2004.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References

We thank Philips Medical Systems for the loan of an ultrasound transducer. We also thank the women of Sydney West Area Health Service for volunteering their time.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Contribution to authorship
  8. Details of ethics approval
  9. Acknowledgements
  10. References