Smoking and Pregnancy

Overview
Tobacco smoking is associated with adverse pregnancy outcomes which may be preventable through smoking cessation interventions. Advice, multiple component programs, behavioral strategies, repeated contacts, and self-help manuals are effective in decreasing tobacco smoking significantly in pregnant women. Interventions are effective in diverse populations with varying levels of nicotine dependence and at different periods of gestation. A reduction in tobacco use increases birth weight, decreases the incidence of low birth weight infants and is cost effective. Cognitive ability is marginally improved in children of mothers who have not smoked during gestation. Further evaluative research is needed on interventions designed to maintain abstinence. Prevention of tobacco-related illnesses in the non-pregnant population is dealt with in Chapter 43.

Burden of Suffering
Smoking during pregnancy harms both the mother and her developing fetus. Aside from increased morbidity and mortality from cancers, cardiovascular and pulmonary disease in the mother, smoking has been implicated in the etiology of abruptio placenta, placenta previa, spontaneous abortion, premature delivery, and stillbirth. Prenatal smoking is thought to account for about 18% of cases of low birth weight (<2500 g), and also increases risk of shortened gestation, respiratory distress syndrome, and sudden infant death syndrome.

Cigarette smoking is the principal cause of low birth weight in developed countries. Intrauterine growth retardation is the most strongly documented adverse effect of smoking during pregnancy. This is a significant public health concern because low birth weight is the most important single determinant of neonatal and infant morbidity and mortality. Retarded fetal growth in the offspring of smokers may be attributable to several factors, including the vasoconstricting properties of nicotine, elevated fetal carboxyhemoglobin and catecholamine levels, fetal tissue hypoxia, reduced delivery of nutritional elements and elevation of heart rate and blood pressure. Even after controlling for alcohol use, socioeconomic status, maternal height, maternal weight and years of education, smoking has been implicated in long-term effects such as poor cognitive performance on achievement tests and decreased physical growth.

In Canada, the incidence of low birth weight in infants of mothers in all age groups declined from 6.6% of 343,000 births in 1971 to 4.6% of 377,00 births in 1989, a 30.3% decline over the 18 year period, comprising mainly birth weights of 1,500 to 2,499 g. The prevalence of birth weights in this range decreased from 5.8% of births in 1971 to 4.0% in 1989, while the prevalence of very low birth weight (<1500 g) remained stable. Most of this decline in low birth weight has been attributed to a decrease in smoking rates in women of reproductive age. The Labour Force Survey Smoking Supplement estimated that smoking rates for Canadian women of reproductive age (15-44 years) declined from 37% in 1972 to 29% in 1986.

Exposure to environmental tobacco smoke (passive smoking) may also have a modest adverse effect on birth weight.<1> Hair concentrations of nicotine and cotinine in women and their newborn infants provide biochemical evidence that infants of smokers and of passive smokers have measurable systemic exposure to cigarette smoke toxins. The clinical significance of this exposure is as yet unclear.

Maneuver
The interventions developed to help pregnant smokers quit that have been evaluated in published research studies include smoking cessation advice, feedback and individual or group counselling.<2> Nicotine replacement therapy has not been adequately studied in pregnant women. Use of such therapy by pregnant women has been advocated by Benowitz<3> because of its benefits as an adjunct to smoking cessation therapy in non-pregnant populations. Nicotine replacement cannot be recommended at present, however, since it could conceivably contribute to adverse effects on the fetus and because its efficacy in pregnant smokers has not yet been established. Interventions aimed at reducing exposure to environmental tobacco smoke have also not been evaluated.

"Smoking Cessation Advice" has been defined as providing health education to tobacco smoking pregnant women to stop smoking.<4-7> The underlying premise has been that if women were aware of the adverse effects of smoking during pregnancy they would stop smoking.<4> Such advice has usually included information about the effects of smoking on the fetus given directly by a physician or midwife, supplemented by a health education booklet. The advantage of this intervention is that it is brief. In the trial reported by Lilley<7> it lasted 10 minutes, and could be given by a physician or midwife, who would ordinarily be in contact with the patient for prenatal care. However, knowledge concerning adverse health effects is necessary but not always sufficient to induce patient compliance.<8> Since addictions are complex behaviours with multifactorial origins, simply giving women information about the ill-effects of smoking and advising them to quit without providing the support needed to achieve that goal may not produce the desired result.

"Feedback" implies evaluating patient status prior to the intervention through a carbon monoxide breath sample, a cotinine blood sample, or a fetal ultrasound. Patients are provided with the results of these measures, sometimes with comparative measures in nonsmoking individuals. Health advice is given about how to improve these measures through smoking cessation.

Multiple component intervention programs combine elements of health education, self-help manuals on how to quit smoking, supportive counselling and multiple follow-up contacts. These interventions are more labour intensive than advice, feedback, or group counselling.

Effectiveness of Treatment
It is estimated that 25% to 40% of pregnant women smokers quit smoking without any intervention for at least a brief time upon learning they are pregnant.

Smoking Cessation Advice
There have been several randomized controlled trials<4-7> of smoking cessation advice among pregnant women. Unfortunately, design problems have included small sample size, poor description of the intervention,<5,6> lack of uniform intervention delivery and contamination of treatment and control groups.<6> Follow-up was reported to be 66% to 100%. Outcomes were based on self-report with only one study<5> reporting biochemical verification. Dropouts were omitted from the final analysis in all studies using advice as the intervention. Quit rates (stopping smoking for the remainder of the pregnancy) were consistently higher (but not statistically higher) in the experimental (6-14%) as opposed to the control groups (1-6%).

A 1993 meta-analysis found that advice significantly reduced the proportion of smokers who continued smoking through pregnancy, compared with smokers who received standard antenatal care (odds ratio 0.39; 95% confidence interval (CI): 0.21-0.75).<9>

In primiparas, MacArthur<6> reported that mean birth weight in the intervention group receiving advice was 68 g heavier than that of controls (p<0.06). The author also noted that primiparas in the intervention group were more likely to have received adequate advice. Sixty one percent of primiparas recalled being advised to stop smoking by the obstetrician or by the midwife, compared with only 45% of multiparas. Mean birth weights of multiparas in the two groups were not statistically different.

Feedback
Three trials have used feedback involving serum cotinine levels, carbon monoxide levels, and ultrasound examinations.<10-12> Blood tests and ultrasound are often already part of antenatal care; testing carbon monoxide levels is non-invasive. Thus, minimal additional cost or time was involved. Although these were randomized trials and provided good descriptions of the interventions, design problems included poor follow-up,<10,12> small subject numbers<10,11> and omission of dropouts from analysis. Again, quit rates were higher (but not statistically higher) in the experimental group. One trial was designed to test a self-reported multifactorial lifestyle change which included drinking and other health-related activities as well as smoking. The number of smokers who reported a change in smoking behaviour was low and the results could have been greatly influenced by omission of drop-outs from the analysis.<11> Thus there is insufficient evidence to evaluate the effectiveness of feedback.

Group Counselling
In a group counselling intervention adapted by Loeb et al<13> from the Multiple Risk Factor Intervention Trial, the significance of the results was limited by the fact that only 10% of the treatment group attended all counselling sessions. Experimental and control groups had similar quit rates (15% and 14%, respectively). Two of three trials<14,15> that compared counselling to usual care found significantly increased abstinence rates after the intervention (14% vs. 8%; and 15% vs. 5%). One small trial of women attending a public health clinic<16> found counselling made no difference (21% vs. 23% abstinence during pregnancy) but the usual intervention was clearly very effective. Most studies that compared post-partum recidivism rates between counselling and control groups found higher relapse rates among quitters in the control groups compared to those in the intervention groups. Group counselling thus has had mixed results and should be evaluated further.

Multiple Component Programs
Several trials have evaluated multiple component programs.<17-22> All but two studies<21,22> were randomized trials, and most had over one hundred subjects with 84-98% follow-up. Except for two studies,<18,19> drop-outs were counted as treatment failures.

Quit rates were significantly increased (p<0.05) by all behavioral strategy interventions (quit rates, experimental groups 10-27%; control groups 2-9%). Clinically significant birth weight differences were observed, with decreased low (<2500 g) and very low birth weight (<1500 g) in infants of those who quit smoking. One study found a 5.6% incidence of low birth weight in the intervention groups compared with 6.52% incidence in the control groups.<19> Ershoff,<19> Gillies<22> and Windsor<23> found smoking cessation interventions were cost effective, comparing the cost of hospital delivery in treated versus control groups including the cost of the intervention. Ershoff found a benefit of 2.8 to 1 for the intervention vs. the control group.

A 1993 meta-analysis of behavioral strategies found a significant reduction in the proportion of smokers who continued smoking through pregnancy, compared with standard antenatal care or with personal advice supplemented by written materials (odds ratio 0.30; 95% CI: 0.23-0.38).<24> However, the author concluded that since even the most effective strategies implemented during pregnancy have a limited effect, obstetricians and midwives should also support population strategies towards progressive reduction in cigarette smoking for society as a whole. In a separate analysis of all interventions to reduce smoking in pregnancy, Lumley concluded that smoking cessation interventions result in a small increase in mean birthweight. Effects on preterm birth and perinatal mortality were unclear.

Maternal Smoking After Pregnancy
Postpartum recidivism was high in studies which included post-intervention< 4,22> and postpartum assessment.<18-21> Sexton found that three years after completing the trial, 72% of those who quit during pregnancy were smoking again and 91% of those who did not quit during pregnancy were still smoking.

Thus, despite having achieved statistically significant quit rates during pregnancy, these gains were not maintained and would not be assumed to improve the mother’s long-term health in the majority of cases. The clinically significant benefit may be limited to the offspring.

Long-term Effects of Maternal Smoking During Pregnancy on Children
Most long-term studies of children whose mothers smoked during pregnancy have focused on growth and neurocognitive ability. Average height and weight of 3-year-old children whose mothers had quit during pregnancy were significantly increased over children of non-quitters (height p<0.001, weight p<0.05).<25> Whether the differences found, (0.45 kg for weight and 1.13 cm for height) were clinically significant may be open to question.

Several cohort and case-control studies have noted differences in psychometric test results in children of women who smoked during pregnancy and children of non-smokers.<26-33> Sexton and colleagues<26> carried out cognitive testing in the three-year-old offspring of mothers who had quit and in children of women who had continued smoking during pregnancy. The Preschool Version of the Minnesota Child Development Inventory and the McCarthy Scales of Children’s Abilities were used as outcome measures. The General Cognitive Index score in children of quitters averaged 5 points higher than in children of non-quitters (p<0.01), even when babies of <2,500 g birth weight were excluded and after controlling for other variables such as socioeconomic status, maternal behaviour, maternal time available to child and child characteristics. Statistically significant differences of one to 3 points were also noted on the McCarthy subscales. McCarthy suggests 15 points between pairs of subscales as a rule of thumb for determining noteworthy differences. Other investigators have reported inconsistent effects of smoking on psychological testing – both 1) significantly lower scores in the smoking group versus the non-smokers and 2) no significant differences between children of smokers and non-smokers. Based on the evidence, one would conclude that smoking during pregnancy may be detrimental to the offspring or at best, smoking has no effect – in no case has smoking been shown to coincide with improved psychometric test scores.

Characteristics of Women Who Quit Smoking During Pregnancy
Four percent of women deny smoking even in the face of biochemical evidence to the contrary. To determine how well women who reported quitting smoking prior to pregnancy were able to maintain that status, several authors have studied "spontaneous quitters" (i.e. women who quit smoking in response to pregnancy before the start of prenatal care). In a randomized controlled trial by Quinn, Mullen, and Ershoff,<19,34> spontaneous quitters were defined as women who stated that they had quit smoking since becoming pregnant and had not smoked for at least 24 hours. This group was compared to a group who reported smoking at least seven cigarettes per week prior to pregnancy. Spontaneous quitters and smokers differed significantly in the following respects: 1) Spontaneous quitters had been lighter smokers prior to pregnancy; 2) they were less likely to have another smoker in their household; 3) indicated a stronger belief in the harmful effect of maternal smoking on fetal health; 4) had a history of fewer miscarriages; and 5) had entered prenatal care earlier. Compared to women who maintained cessation as measured by urine cotinine levels, women who relapsed had less confidence in their ability to stay off cigarettes, were more likely to be multigravidas, and believed less strongly in the harmful effects of maternal smoking on fetal health. Other authors have found that women who quit smoking had higher socioeconomic levels, were older, had smoked fewer cigarettes, and were better educated than women who continued smoking.

Recommendations of Others
The Canadian Nurses Association and the U.S. Preventive Services Task Force<35> recommend that pregnant women receive smoking cessation education. The Canadian Medical Association, American College of Physicians, American College of Obstetricians and Gynecologists, and the American Academy of Pediatrics, recommend that physicians encourage smoking cessation. The Royal College of Physicians and Surgeons of Canada recommend that smokers who wish to stop smoking should receive effective help.

Conclusions and Recommendations
Interventions which include advice, multiple components, behavioral strategies, support, multiple contacts, and self-help manuals are effective in significantly decreasing tobacco smoking in pregnant women. Interventions work with diverse populations with different levels of nicotine dependence and at different stages of gestation. Decrease in tobacco use has a beneficial effect on increasing average birth weight and decreasing the incidence of low birth weight infants. Smoking cessation interventions are cost effective as a result of decreasing the number of low birth weight infants. Cognitive ability is also improved in children of mothers who did not smoke during gestation. Thus, there is good evidence to recommend smoking cessation interventions for all pregnant women who smoke (A Recommendation).

Unanswered Questions (Research Agenda)
More research is needed on interventions to maintain abstinence post-delivery.

Evidence
Information retrieval sources were in consultation with Addiction Research Foundation Library and Fudger Medical Library at Toronto General Hospital using MEDLINE, 1966 to 1993. Key words used include: Smoking, smoking cessation, tobacco; infant, low birth weight, small for gestational age, newborn; birth weight, fetus, growth retardation; abnormalities, brain development; growth, brain growth; psychometrics; child development; pregnancy; prenatal care, exposure, delayed effects; longitudinal studies; evaluation studies. Science Citation Index, 1990-1992: Author’s names in clinical trials.

Expert consultation and review of literature files of: Smoking Cessation Clinic, Community Treatment Research Unit, Addiction Research Foundation Dr. R Frecker. Prevention, Health Promotion, Addiction Research Foundation, M. Pope, and reference sections from articles.

This review was initiated in January 1993 and the recommendations were finalized by the Task Force in June 1993.

Acknowledgements
Funding for this report was provided by Health Canada under the Government of Canada’s Brighter Futures Initiatives. The Task Force thanks Helen P. Batty, MD, CCFP, MEd, FCFP, Associate Professor, Department of Family and Community Medicine, University of Toronto, Toronto, Ontario and Douglas M.C. Wilson, MD, CCFP, FCFP, Professor of Family Medicine, McMaster University, Hamilton, Ontario for reviewing the draft report.

Full Citation
Moner S.E. Smoking and pregnancy. In: Canadian Task Force on the Periodic Health Examination. Canadian Guide to Clinical Preventive Health Care. Ottawa: Health Canada, 1994; 26-36.