E-cigarette use during pregnancy and its association with adverse birth outcomes in the US

Abstract

The popularity of e-cigarette use among young adults is a growing concern. However, little is known about factors associated with e-cigarette use in pregnant women and birth outcomes. In this retrospective cohort study, we evaluated the influence of several factors on behavioral changes in e-cigarette use before and during pregnancy, and assessed the association between e-cigarette use and subsequent birth outcomes among pregnant women. The Population Assessment of Tobacco and Health (PATH) study, a government-sponsored national longitudinal study based in the US, Waves 1 through 4 (2013–2018) were used. Multivariate logistic regressions were conducted to estimate behavioral changes in e-cigarette use during pregnancy and subsequent influence on high-risk birth (e.g., preterm birth, low birth weight, birth defects, etc.) and fetal death. Although pregnant women who quit vaping before pregnancy (OR = 1.14, 95% CI 0.54 – 2.40) or had any use during pregnancy (OR = 1.19, 95% CI 0.38 – 3.73) showed non-differential risk of having a high-risk birth in comparison to women who did not initiate vaping, we observed that the usage of mint/menthol flavor was correlated with higher risk of fetus death (OR = 3.27, 95% CI 1.17 – 9.19). Healthcare providers should encourage e-cigarette users to quit prior to and during early pregnancy.

INTRODUCTION

First introduced into the U.S. market in 2007, electronic cigarettes (e-cigarettes) have quickly become the dominant tobacco product used by young adults. (1) Use rates accelerated between 2014 and 2019 (prevalence of current daily e-cigarette use increased from 5.2% to 9.3%, respectively), but declined afterwards. (2–4) Despite the encouraging decline, it is premature to speak of a trend, since the pandemic could have disrupted longitudinal assessments. (2–4) Of particular concern is women of reproductive age, who are vulnerable to exposure to and use of tobacco products. National estimates showed that the prevalence of current e-cigarette use among pregnant women increased from 2.2% in 2014 to 3.6% in 2018. (5, 6) Previous research indicates that pregnant women tend to vape e-cigarettes and smoke cigarettes concurrently. (2) Concurrent usage is believed to be due to suggestions that e-cigarettes serve as a safer alternative to cigarettes and are potentially conducive to smoking cessation. (7) The growing popularity of e-cigarettes, coupled with the unique health concerns and considerations for pregnant women and the developing fetus, are a major area of public health concern.

Pregnant women and developing fetuses are susceptible to harmful chemicals in tobacco products. Although smoking during pregnancy has been shown to increase health-associated risks for developing babies, including an increased likelihood of preterm birth, low birth weight, and birth defects (8, 9), the association between e-cigarette use and obstetric outcomes is still unclear. One known factor that contributes to the abnormality of a fetus’s development is nicotine (10), and systematic reviews suggest that nicotine consumption in vaping is comparable to smoking.(11) A growing body of experimental studies in animal models have shown that the nicotine present in electronic nicotine delivery systems (ENDS) alters deoxyribonucleic acid (DNA) methylation, which could induce birth defects, reduce birth weight, and distort organ development for the newborn. (12) In addition, the flavorings used within e-cigarettes could be harmful for the developing fetus. (13) Bench research showed that the flavors of e-cigarettes are associated with elevated serum inflammatory cytokines, resulting in DNA damage and cell death (both apoptotic and necrotic). (14) Pregnant mice exposed to e-vapor delivered offspring with deficits in short-term memory and hyperactivity. (14) While these are important findings, limited research has evaluated the associations between e-cigarette use and birth outcomes within a nationally representative cohort of women. (15–17)

From a medical perspective, it is important for various stakeholders to understand factors that can influence behavioral changes, in specific, factors (i.e., social/family influence, psychological coping, reinforcing effects of cigarettes, and stigma) that influence the decision to switch from one tobacco product to another. (18, 19) Knowing these factors can help in designing effective interventions and counseling services to mitigate e-cigarette use during pregnancy. However, research related to the influence of vaping on pregnancy outcomes is still limited. (16, 17)

In this study, we aim to (1) examine factors that may be associated with behavioral changes in e-cigarette use before and during pregnancy; (2) evaluate the association between e-cigarette use before and during pregnancy on adverse birth outcomes (high-risk birth and fetus death; and (3) explore the association between vape flavorings and adverse birth outcomes (high-risk birth and fetus death).

MATERIALS AND METHODS

Data and Study Design

Data from the Population Assessment of Tobacco and Health (PATH) study was used to address the aims of this study. (20) The PATH study, which is conducted across the U.S., constitutes a national sample of tobacco users and non-users. Data collection for PATH started in 2011with a cohort of 45,971 individuals (baseline or Wave 1). The Wave 4 ran from December 1, 2016, to January 3, 2018, and consisted of 33,822 adults (25,856 continuing adults and 6,065 new cohort adults). (20) Details on the design and methodology of the PATH study have been described previously (21). The PATH public file was determined by the George Mason University IRB to be exempt as non-human research.

This present study is a retrospective cohort study that utilized all available data: PATH Waves 1–4. The study included women of reproductive age (18 to 44 years old) who were pregnant between Waves 1 and 3 and delivered during the subsequent survey round.

Sample Selection

The PATH consisted of 23,301 women of reproductive age (18–44 years old). Women who were in their pregnancy or encountered labor within one year of the interview date were included in this study (n=2,161). Those excluded were male respondents, women who were not in the reproductive-age group (> 44 years old),(22) and women who had missing data on vaping or smoking status before or during their pregnancy (See supplementary Figure S1 for more detail). In this sample, 1,130 women were included to examine factors influencing changes in e-cigarette use before and during pregnancy (non-user n=728, quit before pregnancy n=265, quit during pregnancy n=96, and continue vaping n=41). We included 597 women who had birth information available to evaluate the association between e-cigarette use and high-risk birth and fetus death (high-risk birth n=75, fetus death n =34, normal birth n=488).

Measures

Outcome measures

One of the outcome variables was women’s behavioral change related to e-cigarette use during their pregnancy. This outcome variable consisted of four mutually exclusive groups, namely, the “never users,” “quit before pregnancy,” “quit during pregnancy,” and “continuous users.” In the PATH study, women who did not currently use e-cigarettes or had not used in the past 12 months were asked to report the length of time since their last use of an e-cigarette product. Pregnant women also provided information on the duration of their pregnancy. We combined the information on e-cigarette use status and length of pregnancy to determine whether a woman had quit e-cigarettes before or during pregnancy. Continuous users and never users were defined by questions asking about their vaping status. More details regarding the construction of outcomes can be found in Supplements S2.

In evaluating the association between e-cigarette use before and during pregnancy and adverse birth outcomes, the main outcome variables were high-risk birth and fetal death. (23) High-risk birth was defined as any of the following that occurred in the last pregnancy: preterm birth, low birth weight, baby with birth defects, placenta previa, placenta abruption, pre-eclampsia, and cleft lip or palate. (20, 24) Fetal death was defined as having a miscarriage, abortion, or ectopic or tubal pregnancy. (20)

Main independent measures

In examining factors that influence changes in e-cigarette use before and during pregnancy, we utilized previous research findings¹⁵ to classify them into three groups: social factors, personal belief and health status, and cigarette smoking status. Social factors consisted of educational attainment, race, maternal age, and level of satisfaction with social activities and relationships. Personal belief consisted of an individual’s general perception of e-cigarettes compared to other tobacco products; whether the individual received advice from a healthcare professional to quit tobacco/e-cigarettes within the prior 12 months (before or during the pregnancy); an individual’s overall health status (self-reported), including both physical and mental health; and an individuals’ cigarette, alcohol, and marijuana use prior to and during pregnancy, and whether smoking is allowed at home.

In examining the association between e-cigarette use and birth outcomes, the behavioral change measure related to e-cigarette use served as the primary independent variable. As described above, this consists of not using e-cigarettes (non-user), quit before pregnancy, and any usage of e-cigarettes during pregnancy (vaping throughout pregnancy [continued vaping] or quit during pregnancy). In exploring the relationship between e-cigarette flavors and adverse birth outcomes, common flavors of e-cigarettes were included such as mint, alcohol, and candy (25, 26). Each flavor is a dummy variable where 1 represented marked and 0 was other flavors.

Covariates

In examining the association between e-cigarette use and birth outcomes, same social factors, personal beliefs and health, and smoking status stated above were included as covariates. In exploring the relationship between e-cigarette flavors and adverse birth outcomes, we built the regression models in two-steps. First, we conducted the univariate analysis including only the e-cigarette flavors (mint/menthol, alcohol, candy, and others). Second, same covariates used in the first two aims were included.(27)

Statistical Analysis

Descriptive analyses were conducted to assess distribution of high-risk birth and fetus death using weighted estimates. Unweighted logistic regressions were fitted to estimate the association between factors and behavior changes of e-cigarettes use during pregnancy, as well as the associations between e-cigarette use and birth outcomes. Multiple imputation and further analysis was conducted as a sensitivity test to account for the missing values. More details can be found in Supplement S3. All statistical analyses were conducted using SAS version 9.4 (SAS institute INC, Cary, NC) and STATA 16.1 (StataCorp LP, College Station, TX).

RESULTS

Factors Associated with Quitting e-Cigarettes before and during Pregnancy

Table 1 shows the factors associated with behavioral changes in e-cigarette use for pregnant women. Compared to those who had quit vaping before pregnancy, women who did not smoke were also more likely not to use e-cigarettes (OR = 2.32 95% CI 1.04 – 5.21, p < 0.05). In contrast, women who smoked in pregnancy—including those who opted to quit during pregnancy—were more likely to vape during pregnancy as well (OR = 11.39 95% CI 1.36 – 95.33, p < 0.05). We noticed that, compared to women who were racially White, African Americans (OR = 3.77, 95% CI 2.10 – 6.78, p < 0.01) were less likely to initiate vaping. We also noticed that having a harmful perception of e-cigarette use was a significant predictor for not using e-cigarettes during pregnancy. Women who regarded e-cigarette use to be as harmful as cigarette use had a significantly higher chance of being non-users (OR = 1.72, 95% CI 1.152 – 2.57, p < 0.01) and were less likely to vape throughout pregnancy (OR = 0.44, 95% CI 0.25 – 0.79, p < 0.01). Further, those who considered e-cigarette use more harmful than cigarette use showed higher odds of not initiating e-cigarette use (OR = 2.23, 95% CI 1.00 – 4.97, p < 0.05).

Table 1:

Factors Associated with Behavioral Changes in e-Cigarette Use during Pregnancy

	E-cigarette use
	Non-user vs Quit before pregnancy	Any use during pregnancy vs Quit before pregnancy
	AORa (95% CI)	AORa (95% CI)
Cigarette use
Quit before pregnancy	Ref.	Ref.
Not using cigarette	2.32**	5.51
	(1.04 – 5.21)	(0.67 – 45.16)
Any use during pregnancy	0.92	11.39**
	(0.37 – 2.27)	(1.36 – 95.33)
Age
18 to 24 years old	Ref.	Ref.
25 to 34 years old	1.47	1.70
	(0.98 – 2.19)	(0.94 – 3.07)
35 to 44 years old	1.79	1.63
	(0.91 – 3.51)	(0.61 – 4.34)
Education level
Less than high school	Ref.	Ref.
GED	0.64	0.88
	(0.28 – 1.49)	(0.26 – 2.97)
High school graduate	0.90	1.07
	(0.47 – 1.73)	(0.41 – 2.78)
Some college (no degree) or associates degree	0.56	0.73
	(0.30 – 1.04)	(0.29 – 1.80)
Bachelor’s or advanced degree	2.03	0.57
	(0.94 – 4.40)	(0.16 – 1.97)
Race
White	Ref.	Ref.
African American	3.77***	0.86
	(2.10 – 6.78)	(0.36 – 2.05)
Others	1.16	0.44
	(0.67 – 1.99)	(0.15 – 1.27)
Physical health before pregnancy
Very good/Excellent	Ref.	Ref.
Good	1.02	0.92
	(0.66 – 1.57)	(0.48 – 1.78)
Fair/Poor	1.53	0.75
Mental health before pregnancy	(0.80 – 2.93)	(0.30 – 1.86)
Very good/Excellent	Ref.	Ref.
Good	0.94	1.29
	(0.60 – 1.46)	(0.64 – 2.58)
Fair/Poor	0.62	1.50
	(0.36 – 1.06)	(0.70 – 3.19)
Level of satisfaction with social activities and relationships
Very satisfied	Ref.	Ref.
Moderately satisfied	0.72	1.06
	(0.46 – 1.11)	(0.57 – 1.99)
Little or not satisfied	1.04	0.84
	(0.49 – 2.21)	(0.27 – 2.59)
Harmful perception of e-cigarettes compared to cigarettes
Less harmful	Ref.	Ref.
About the same	1 72***	0.44***
	(1.15 – 2.57)	(0.25 – 0.79)
More harmful	2.23**	0.74
Smoking allowed at home	(1.00 – 4.97)	(0.21 – 2.62)
Not allowed	1.13	0.69
	(0.69 – 1.84)	(0.32 – 1.46)
Received advice to quit tobacco products in pregnancy	0.61**	1.44
	(0.38 – 0.97)	(0.79 – 2.63)
Alcohol use
Quit before pregnancy	Ref.	Ref.
Not using	1.32	1.06
	(0.76 – 2.27)	(0.55 – 2.06)
Any use during pregnancy	1.96**	0.77
	(1.09 – 3.50)	(0.34 – 1.72)
Marijuana use
Quit before pregnancy	Ref.	-
Not using	2.16** (1.03 – 4.55)
Any use during pregnancy	1.28
	(0.59 – 2.74)
Observations	751	304

Notes:

a: OR=Adjusted Odds Ratio. 95% Confidence Intervals were in the parenthesis. - means not applicable due to the exclusion of insufficient observations

^***p < 0.01

^**p< 0.05

The results show that women who reported any alcohol consumption in pregnancy, compared to those who stopped drink before pregnancy, were more likely not not use e-cigarettes (OR = 1.96, 95% CI 1.09 – 3.50, p < 0.05). On the other hand, women who did not use marijuana in their lifetime had higher odds of not using e-cigarettes, either (OR = 2.16, 95% CI 1.03 – 4.55, p < 0.05).

Associations between Birth Outcomes and Smoking/Vaping Behavior

Table 2 presents the associations between behavioral changes in e-cigarette use and subsequent birth outcomes. We found that, although having a history of e-cigarette use—whether or not to quit before pregnancy or use during pregnancy—was associated with a higher risk of a high-risk birth (OR = 1.14 – 1.19), the effects did not reach statistical significance (95% CI ranged from 0.38 to 3.73). In regard to associated factors of health, we found that African American women were more likely to experience a fetus death compared to their White counterparts (OR = 2.74, 95% CI 1.03–7.26, p < 0.05).

Table 2:

Associations between e-Cigarette Usage and Adverse Birth Outcomes

	(1)	(2)
	High-risk birtd N = 375	Fetus death N = 402
	AORa (95% CI)	AORa (95% CI)
E-cigarette use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	1.14	0.39
	(0.54 – 2.40)	(0.07 – 2.26)
Any use during pregnancy	1.19	0.33
	(0.38 – 3.73)	(0.04 – 2.83)
Cigarette use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	2.69*	0.95
	(0.97 – 7.42)	(0.18 – 4.91)
Any use during pregnancy	1.51	0.56
	(0.62 – 3.68)	(0.15 – 2.15)
DEMOGRAPHIC FACTORS
Age
18 to 24 years old	Ref.	Ref.
25 to 34 years old	1.24	0.81
	(0.60 – 2.57)	(0.31 – 2.12)
35 to 44 years old	1.32	0.57
	(0.52 – 3.35)	(0.10 – 3.17)
Education level
Less than High School	Ref.	Ref.
GED	0.74	0.46
	(0.15 – 3.74)	(0.06 – 3.34)
High school graduate	0.55	0.08***
	(0.17 – 1.79)	(0.02 – 0.43)
Some college (no degree) or associates degree	1.32	0.36
	(0.45 – 3.84)	(0.09 – 1.39)
Bachelor’s or advanced degree	1.53	0.10**
	(0.43 – 5.41)	(0.01 – 0.80)
Race
White	Ref.	Ref.
African American	1.61	2.74**
	(0.69 – 3.74)	(1.03 – 7.26)
Others	0.75	0.22
	(0.28 – 2.00)	(0.02 – 1.98)
Level of satisfaction with social activities and relationships
Very/Extremely satisfied	Ref.	Ref.
Moderately satisfied	0.32**	1.70
	(0.12 – 0.82)	(0.57 – 5.04)
A little/Not satisfied	0.66	1.16
	(0.18 – 2.36)	(0.17 – 7.68)
General perception of e-cig to cigarettes
Less harmful	Ref.	Ref.
About the same	0.84	1.02
	(0.43 – 1.65)	(0.33 – 3.12)
More harmful	1.19	2.03
	(0.35 – 4.00)	(0.30 – 13.80)
e-Cigs allowed at home [Yes or no, here? Unclear!]	0.86	0.87
	(0.35 – 2.07)	(0.26 – 2.95)
Received advise to quit tobacco/e-cigarettes in pregnancy	0.85	1.80
	(0.36 – 2.00)	(0.59 – 5.54)
Physical health before pregnancy
Very good/excellent	Ref.	Ref.
Good	1.87	1.24
	(0.90 – 3.88)	(0.43 – 3.61)
Poor/fair	2.01	3.51
	(0.63 – 6.43)	(0.69 – 17.82)
Mental health before pregnancy
Very good/excellent	Ref.	Ref.
Good	0.93	1.92
	(0.43 – 2.00)	(0.64 – 5.74)
Poor/fair	1.61	0.96
	(0.62 – 4.18)	(0.22 – 4.15)
Alcohol use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	0.69	1.32
	(0.29 – 1.64)	(0.37 – 4.63)
Any use during pregnancy	0.54	0.47
	(0.23 – 1.23)	(0.13 – 1.66)
Marijuana use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	0.58	1.17
	(0.19 – 1.73)	(0.30 – 4.53)
Any use during pregnancy	0.93	0.11**
	(0.43 – 2.02)	(0.01 – 0.95)
	AORa (95% CI)	AORa (95% CI)
E-cigarette use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	1.14	0.39
	(0.54 – 2.40)	(0.07 – 2.26)
Any use during pregnancy	1.19	0.33
	(0.38 – 3.73)	(0.04 – 2.83)
Cigarette use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	2.69*	0.95
	(0.97 – 7.42)	(0.18 – 4.91)
Any use during pregnancy	1.51	0.56
	(0.62 – 3.68)	(0.15 – 2.15)
Demographic factors
Age
18 to 24 years old	Ref.	Ref.
25 to 34 years old	1.24	0.81
	(0.60 – 2.57)	(0.31 – 2.12)
35 to 44 years old	1.32	0.57
	(0.52 – 3.35)	(0.10 – 3.17)
Education level
Less than High School	Ref.	Ref.
GED	0.74	0.46
	(0.15 – 3.74)	(0.06 – 3.34)
High school graduate	0.55	0.08***
	(0.17 – 1.79)	(0.02 – 0.43)
Some college (no degree) or associates degree	1.32	0.36
	(0.45 – 3.84)	(0.09 – 1.39)
Bachelor’s or advanced degree	1.53	0.10**
	(0.43 – 5.41)	(0.01 – 0.80)
Race
White	Ref.	Ref.
African American	1.61	2.74**
	(0.69 – 3.74)	(1.03 – 7.26)
Others	0.75	0.22
	(0.28 – 2.00)	(0.02 – 1.98)
Level of satisfaction with social activities and relationships
Very/Extremely satisfied	Ref.	Ref.
Moderately satisfied	0.32**	1.70
	(0.12 – 0.82)	(0.57 – 5.04)
A little/Not satisfied	0.66	1.16
	(0.18 – 2.36)	(0.17 – 7.68)
General perception of e-cig to cigarettes
Less harmful	Ref.	Ref.
About the same	0.84	1.02
	(0.43 – 1.65)	(0.33 – 3.12)
More harmful	1.19	2.03
	(0.35 – 4.00)	(0.30 – 13.80)
e-cigs allowed at home	0.86	0.87
	(0.35 – 2.07)	(0.26 – 2.95)
Received advise to quit tobacco/e-cigarettes in pregnancy	0.85	1.80
	(0.36 – 2.00)	(0.59 – 5.54)
Physical health before pregnancy
Very good/excellent	Ref.	Ref.
good	1.87	1.24
	(0.90 – 3.88)	(0.43 – 3.61)
Poor/fair	2.01	3.51
	(0.63 – 6.43)	(0.69 – 17.82)
Mental health before pregnancy
Very good/excellent	Ref.	Ref.
good	0.93	1.92
	(0.43 – 2.00)	(0.64 – 5.74)
Poor/fair	1.61	0.96
	(0.62 – 4.18)	(0.22 – 4.15)
Alcohol use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	0.69	1.32
	(0.29 – 1.64)	(0.37 – 4.63)
Any use during pregnancy	0.54	0.47
	(0.23 – 1.23)	(0.13 – 1.66)
Marijuana use
Not using e-cig	Ref.	Ref.
Quit before pregnancy	0.58	1.17
	(0.19 – 1.73)	(0.30 – 4.53)
Any use during pregnancy	0.93	0.11**
	(0.43 – 2.02)	(0.01 – 0.95)

Notes:

^a:Adjusted Odds Ratio; -indicates variables not estimable in the regression model. The models further controlled the interactions between cigarette and e-cigarette usage; and e-cigarette usage and race. Standard errors in parentheses.

^**p<0.05

Given the growing concern about the adverse effect of various e-cigarette flavors on the growing fetus, we further investigated their relationships with birth outcomes (shown in Table 3). We found that the use of menthol or mint before and during pregnancy, compared to other flavors, was associated with two times higher of odds of fetal death (OR = 3.27, 95% CI 1.17 – 9.19, p < 0.05). Although not statistically significant, a similar effect of elevated risk for high-risk birth and fetal death was also observed in vaping candy flavors (Table 3).

Table 3:

The Impact of e-Cigarette Flavors on Adverse Birth Outcomes

E-cigarette flavors	(1)	(2)£	(3)	(4) £
	High-risk birth n=113	High-risk birth n=76	Fetus death n=193	Fetus death n=132
	Crude model	Full model	Crude model	Full model
	ORa (95% CI)	ORa (95% CI)	ORa (95% CI)	ORa (95% CI)
Flavor usually/last used before or during pregnancy
Menthol/Mint vs other	0.73	1.09	1.81	3.27**
	(0.28 – 1.90)	(0.22 – 5.49)	(0.94 – 3.50)	(1.17 – 9.19)
Alcohol vs other	0.23	-	0.15**	0.07**
	(0.01 – 4.11)		(0.03 – 0.76)	(0.01 – 0.85)
Candy vs other	1.49	1.89	0.72	1.27
	(0.51 – 4.33)	(0.38 – 9.25)	(0.39 – 1.34)	(0.53 – 3.07)

Notes:

^£Model (2) and (4) controlled for the age, education level, race, level of satisfaction with social activities and relationships, perception of e-cigarette, physical and mental health before and during pregnancy. Cigarette, e-cigarette, alcohol, and marijuana use were excluded due to high collinearity with e-cigarette flavors.

DISCUSSION

Principal Findings

In this study, although we found that pregnant women who had quit vaping before pregnancy or had any use during pregnancy experienced a higher risk of having a high-risk birth in comparison to women who did not initiate vaping, the effects did not reach statistical significance. The exploratory results suggested, however, that vaping with mint or menthol flavor was correlated with a statistically significant higher risk of fetal death.

National and statewide e-cigarette use rates range from 3.6% to 7% (28, 29), with more than 70% of women quitting e-cigarettes during pregnancy (29). Our findings from the multiple imputation show that 24.73% of women had initiated e-cigarette use before pregnancy; of those, 66.96% ‘quit before pregnancy,’ 24.38% ‘quit during pregnancy,’ and less than 9% were ‘continuous users’ throughout their pregnancy. Although it is encouraging that more women opted to quit e-cigarette use after becoming pregnant, there were still women who continued using e-cigarettes during pregnancy. This fact requires more research to determine factors that are associated with this harmful behavior.

Consistent with previous literature (30, 31), we found that women who were African American or other racial groups were less likely to initiate vaping in comparison with the white population. Evidence has shown that white smokers were more likely to transition to e-cigarette use, in part attributable to social norms and awareness of the harmfulness of e-cigarettes compared to cigarettes (30). Our findings suggest that women who considered e-cigarettes as harmful as cigarettes were less likely to initiate e-cigarette use and more likely to quit e-cigarette use during pregnancy. Many factors may shape these attitudes and beliefs. Several studies have observed circumstances where pregnant women, thought to be using e-cigarettes, were subject to harassment, criticism, and stigma during web-based online forums (32, 33). This, coupled with a lack of knowledge about and research into the safety of e-cigarette use, may make women more skeptical of their use (32, 33).

The existing public health concerns about the adverse health effects of e-cigarette use on children or the mothers themselves is consistent with our findings (33). Evidence has suggested that women who planned to be pregnant in the near future considered the fact that maternal e-cigarette use can lead to stillbirth, problems with childhood anxiety, and later diagnoses of ADHD, than those who were currently pregnant (34). With more people sharing health-related information on social media (35), the use of these venues to propagate important health messages can be a potential strategy in helping women who might be considering pregnancy or are currently pregnant to choose to quit smoking.

Other kinds of high-risk behaviors are associated with e-cigarette use in pregnancy. Consistent with the existing literature, we observed that women who did not use marijuana were also less likely to use e-cigarettes during pregnancy. (6) Interestingly, however, our results suggested that women who stopped drinking alcohol before pregnancy were more likely to use e-cigarettes than those who consumed alcohol during pregnancy. Although evidence indicated that pregnant smokers were more likely to switch to e-cigarettes, (36) no evidence has so far been observed for alcohol usage. It is possible that because women consider e-cigarettes as a safter alternative to other substance use (i.e., conventional smoking and marijuana), they switch from alcohol drinking to e-cigarettes in an attempt to reduce the risk of adverse birth outcomes.

Clinical Implications

Our findings show that e-cigarette use was not significantly associated with adverse birth outcomes. Although strong evidence shows adverse effects of maternal cigarette smoking on birth outcomes, such as preterm birth (37), evidence on the potential harmful effects of e-cigarette use is still scarce. Studies conducted in vitro have provided some direct evidence that maternal e-cigarette exposure is detrimental to the renal development of the offspring (38). Research evidence also suggests that dual-users (using e-cigarettes and cigarettes concurrently) in late pregnancy were estimated to have elevated risk of small-for-gestational-age (SGA) neonates but a similar risk of preterm birth compared to non-users (39). However, when compared to cigarette users, the adverse birth outcomes of e-cigarette users were not statistically different (40).

Although our main finding is consistent with current literature, we add to and extend the current line of research by showing that e-cigarette flavors, in particular mint/menthol, were correlated with a higher risk of fetal death. Our results suggest that use of flavored e-cigarettes during pregnancy could pose a higher risk for fetal development compared to other vaping flavors. With the emerging evidence about the influence of nicotine and other toxins on the development of the fetus (41), professional mental and behavioral support provided by healthcare practitioners before, during, and after pregnancy could be an important tool to spark behavioral changes and improve birth outcomes in this group of women.

Policy Implications

With growing concerns about various e-cigarette flavors adversely affecting birth outcomes (41), the Food and Drug Administration (FDA) has proposed rules prohibiting menthol cigarettes in 2022 (42). Our findings reinforce the action of such US tobacco control policies by showing that the usage of popular menthol and mint flavors was associated with a greater increase in the risk of fetal death. The current literature on the choice of flavored e-cigarettes on birth outcomes is very limited because of the novelty of this tobacco product. Future in vitro and in utero research is urgently needed to assess the mechanism(s) and influence of flavors in e-cigarette aerosols on birth outcomes.

Strengths and Limitations

This is one of the first observational cohort studies investigating the association between e-cigarette use and adverse birth and pregnancy outcomes. We utilized a U.S. nationally representative sample with comprehensive questions regarding tobacco use. However, the study has some limitations. First, the analytical sample size of pregnant women who vaped e-cigarettes was small, which reflects the low prevalence of e-cigarette use during pregnancy. Small sample sizes can yield less consistent estimates. In addition, missing data indicates the potential of selection bias and under-reporting where pregnant women might skip some questions for various reasons. For instance, very few pregnant women responded to what e-cigarette flavors they chose to vape. To address this issue, we performed multiple imputation by chained equation as a sensitivity test. This test confirmed the robustness of our results. Second, the types of e-cigarette devices have evolved over time, and the present study did not stratify use by brands due to the limited sample size. Finally, the estimated effects of e-cigarette use on adverse birth and pregnancy outcomes in the absence of medical history about the respondents might be biased because women who have chronic conditions (such as hypertension and diabetes) are more likely to have adverse birth outcomes. Although few women responded to questions such as whether they had been diagnosed with hypertension or diabetes, we included physical health as a control to capture the influence of some chronic conditions.

CONCLUSION

Pregnant women who vaped mint-/menthol-flavored e-cigarettes during pregnancy showed elevated risk of experiencing fetal death than those who vaped other flavors. However, those who opted to quit vaping before pregnancy showed no differences in the likelihood of having a high risk birth compared to non-users. More research with a longer follow-up period is needed to examine the mechanisms behind behavioral choices related to e-cigarette use and flavor selections and their impacts on adverse pregnancy and birth outcomes. Healthcare providers should encourage e-cigarette users to quit prior to and during pregnancy.

Highlights.

• Quitting vaping before pregnancy was not associated with high-risk birth

• Vaping during pregnancy was not associated with high-risk birth

• The usage of mint/menthol flavor was correlated with higher risk of fetus death