Background
Preterm birth (< 37 weeks gestational age) and low birth weight (< 2500 g) are ongoing issues globally, with rates increasing in many countries, despite our advances in knowledge and interventions aimed at known risk factors [
1,
2]. In Canada, Alberta’s provincial preterm birth (9.2% 2019/2020) [
3] and low birth weight (7.5%, 2019/2020) [
3] rates are among the highest in the country, though the reasons for this are not immediately clear. Both preterm birth and low birth weight have been associated with increased risk of cardiovascular disease and diabetes in adulthood [
4,
5].
Preterm birth and low birth weight have multifactorial etiologies, likely affected by a combination of genetic, biological, behavioral, environmental, social, and other factors [
6,
7]. Among these are exposure to prenatal stress and tobacco smoke. There is evidence to suggest a role of for maternal stress in terms of adverse birth outcomes [
8‐
10]. For example, Shapiro et al. [
11] observed mostly consistent associations between perceived stress, pregnancy-related anxiety, and increased the risk of preterm birth, a finding also echoed in other reviews [
9,
10]. Similarly, the risks associated with tobacco smoke exposure (active maternal and environmental tobacco smoke) during pregnancy have been well documented [
12‐
15]. Smoke exposure has been linked to both preterm birth and low birth weight [
12,
16,
17].
There is evidence to indicate that these independent relationships may be modified by other factors [
7,
18‐
22]. Specifically, stress and smoke exposure may interact through pathways such as decreased uterine-placental blood flow, vasoconstriction of placental vessels or changes in immunologic functioning [
23‐
25] Active maternal smoking may also be related to maternal stress. Maternal psychological stress was observed to be higher among pregnant women who smoked in a prospective study in Brazil [
20]. Smoking may be a coping mechanism for pregnant women experiencing prenatal anxiety [
26,
27] Stress during pregnancy could also act as a smoking cessation barrier [
28,
29]. Data from the United States showed that though up to 45% of women are prompted to quit smoking during pregnancy, about 10% continue to smoke [
29]. However, to date, little has been explored regarding potential interactions between stress and smoke exposure and its impact in terms of preterm birth and low birth weight risk. Disparities exist in terms of the population distribution of adverse birth outcomes that might be more fully explained by joint effects [
11,
30]. Stress may lead to an increase in maternal smoking. Alternatively, smoking may also lower women’s perceived stress levels. Clinically, findings related to interactive effects could be important for designing more effective interventions that include a combination of targeted approaches (e.g. smoking cessation plus stress-reduction). In this study, we aimed to investigate the joint effects of total smoke exposure (active maternal and environmental tobacco smoke) and maternal stress (perceived stress and anxiety) on the risk of preterm birth and low birth weight and potential relationships between active maternal smoking and maternal stress among pregnant women in the All Our Families cohort.
Discussion
In this study, there was no evidence of interactive effects between total smoke exposure and measures of perinatal stress in relation to preterm birth and low birth weight. With imputation, a marginal association between the interaction of active maternal smoking and STAI scores and low birth weight was observed. A positive association was also found between active maternal smoking and STAI scores. This may indicate that certain measures of perinatal stress are linked to the propensity to smoke during pregnancy.
Work is emerging regarding behavioral-psychosocial interactions such as smoking and stress or social support in pregnant women. Lobel et al. [
27] found that pregnancy-related stress predicted cigarette smoking, a finding consistent with those in this study. Findings also showed both a direct and indirect (via smoking) relationship with low birth weight [
27]. Eisenbruch et al. also observed that among women who experienced low social support, a greater proportion of women reported smoking compared to those receiving high social support (34% vs. 17%, respectively) [
18]. Another study from Germany also observed that smokers with low social support were 3.3 times more likely to have a pregnancy complication compared to smokers with high social support [
18]. A recent review also observed evidence from several studies that supported an association between perceived stress or number of identified stressors and smoking during pregnancy [
43].
Stress during pregnancy may act as a smoking cessation barrier [
28,
29]. Bullock et al. [
44] investigated the differences between nonsmoking women, women who were successful in quitting smoking during pregnancy, and those who were unable to quit. Findings indicated that differences were observed between the groups in terms of psychosocial stressors, including financial worries, lack of support, and domestic violence [
44]. Women who continued smoking during pregnancy were found to have higher levels of stress, lower levels of social support, and were more likely to experience domestic abuse [
44]. Thus, the inability to quit smoking during pregnancy may be related to factors beyond just addiction and knowledge deficits [
44].
Another hypothesis could be that smoking is a coping mechanism for pregnant women experiencing pregnancy-related anxiety. However, smoking may actually worsen negative emotional states such as stress [
45‐
48]. The perceived relief experienced after smoking may be partially attributable to the alleviation of nicotine withdrawal symptoms rather than a reduction in stress levels [
49,
50]. In a U.S.-based randomized trial evaluating different smoking cessation and postpartum relapse approaches, lower perceived stress levels were associated with smoking cessation in early pregnancy but not late pregnancy, [
28] which could also have ramifications for the timing of interventions.
Prior studies have observed the modification of the stress-preterm birth/low birth weight relationship by other exposures. Nkansah-Amankra et al. [
19] observed this relationship was modified by neighborhood context, with those living in deprived neighborhoods having increased risks of preterm birth and low birth weight. Social environments may affect the risk of preterm birth and low birth weight through stress-related pathways [
51]. Similarly, stress and smoke exposure mechanisms may overlap through shared pathways, affecting the risk of adverse birth outcomes. Maternal stress, nicotine, and carbon monoxide in blood may affect fetal and placental development as well as decrease blood flow between the uterus and placenta [
23,
52‐
54]. Both carbon monoxide and stress may also act as a vasoconstrictor of placental blood vessels [
23,
52,
53]. Smoke and stress exposure may also affect immunologic functioning and inflammatory responses, leading to adverse birth outcomes [
6,
25,
52,
55].
Study findings also suggest that stress, as assessed through Perceived Stress Scale and Spielberger State-Trait Anxiety Inventory scores, are independently associated with an increased odds of preterm birth and/or low birth weight. Both active maternal and total smoke exposure were also observed to be independently associated with preterm birth but not low birth weight in this study. Interactive effects between active maternal smoking, total smoke exposure (active maternal and environmental) and maternal stress were not observed, though active smoking and maternal stress were positively associated.
Our findings were also consistent with other studies that observed an independent positive association between higher maternal stress and preterm birth and low birth weight. Bussières et al. [
10] observed in their meta-analysis of prospective studies that prenatal stress was modestly associated with decreased birth weight and shorter gestational age. Very preterm birth and extremely low birth weight exacerbate the risk of neurobehavioral other impairments in children [
56,
57].
Data from the 2006 Canadian Maternal Experiences Survey show an estimated prevalence of 10.5% for smoking during pregnancy across Canada, with a reported prevalence of 11.8% in Alberta, [
58] which is consistent with the prevalence of smoking in this study (11.6%). Several studies have shown both maternal and environmental tobacco total smoke exposure to be a risk factor for adverse birth outcomes [
12‐
17,
59]. In this study, we found a significant association between both active maternal smoking, total smoke exposure and preterm birth but not low birth weight. In utero smoke exposure (active maternal, environmental tobacco smoke) has also been linked to an increased risk of early adult-onset diabetes and childhood overweight and obesity [
60‐
62].
Limitations
There are several limitations of the study that must be kept in mind when interpreting these findings. One limitation was the lack of a more refined measure of smoking, which is a recommendation for future work in this area. Information about smoking was also only collected at baseline (< 25 weeks gestation) so smoking habits or exposures may have changed by the time of the second wave of data collection (34–36 weeks gestation). However, prior studies have shown that those who report smoking at the beginning of their pregnancy are likely to continue throughout, though fluctuations in smoking intensity were reported as occurring [
63,
64]. Further, the cohort is predominantly white/Caucasian women, which may limit the generalization of these findings to other ethnic groups. Another limitation is the possibility of residual confounding. Sample size and missing data were also other limitations of this study. However, we used a robust method of imputation (with sensitivity analyses) to address this issue. A stress questionnaire specific to pregnant women was not used, though both the PSS and STAI have been used and validated in several other studies [
65‐
70]. As the mechanisms by which stress and smoke exposure may lead to adverse birth outcomes is unknown, the examination of small-for-gestational age is also a potentially important outcome to assess [
71,
72].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.