Biomarkers of Toxicant Exposure and Inflammation Among Women of Reproductive Age Who Use Electronic or Conventional Cigarettes

Abstract

Background: Electronic cigarettes (e-cigarettes) generally have a more favorable toxicant profile than conventional cigarettes; however, limited information exists for women of reproductive age (WRA). Our aim was to compare biomarkers of toxicant exposure, inflammation, and oxidative stress among WRA who self-report exclusive e-cigarette use, exclusive cigarette smoking, or never tobacco use (controls).

Methods: Multivariable linear regression models were used to compare the geometric means of urinary biomarkers of toxicant exposure and their metabolites, serum markers of inflammation [highly sensitive C-reactive protein, soluble intercellular adhesion molecule (sICAM), interleukin 6, fibrinogen], and a measurement of oxidative stress [prostaglandin F2a-8-isoprostane (F2PG2a)] among WRA from the Population Assessment of Tobacco and Health survey.

Results: E-cigarette users had higher levels of lead, tobacco-specific nitrosamines, nicotine metabolites, and some volatile organic compounds (VOCs) than controls. Except for cadmium and lead, e-cigarette users had lower levels of the analyzed urinary toxicant biomarkers compared with cigarette smokers. Cigarette smokers had higher levels of all the biomarkers of toxicant exposure than controls. There were no significant differences in the levels of markers of inflammation and oxidative stress between e-cigarette users and controls. E-cigarette users and controls had lower levels of sICAM and F2PG2a than cigarette smokers.

Conclusion: WRA who use e-cigarettes had lower levels of some of the evaluated urinary biomarkers of toxicant exposure and serum biomarkers of inflammation and oxidative stress than those who smoke cigarettes, but higher lead, nicotine metabolites, and some VOCs than controls, which can increase health risks.

Background

Smoking adversely affects the health of both men and women, including increasing the risks of various cancers, cardiovascular disease, lung disease such as chronic obstructive lung disease, and other health risks.^1,2 Some health risks are unique to female smokers, or women are at higher risk than men for a similar level of exposure. Consequently, tobacco use is a particularly serious health threat for women of reproductive age (WRA).³ Cigarette smoking exposes smokers to >6,000 chemicals and 69 known carcinogens.¹ Cigarette smoking decreases fertility,⁴ and increases the risk of pregnancy complications,^3,5–8 cervical cancer, early menopause, and other health conditions.⁹ Although the prevalence of smoking is decreasing among WRA, >20% of U.S. women age 15–44 years are smokers,¹⁰ and ∼14% of women report smoking during pregnancy.^11,12

The adverse health outcomes of cigarette smoking have been attributed to exposure to nicotine, carbon monoxide, carcinogens, oxidant gases, and heavy metals as well as other chemicals present in tobacco smoke.³ Although smoking adversely affects the health of both men and women, some health risks are unique to female smokers. For reproductive effects, nicotine (the addictive substance in tobacco products), cadmium (a heavy metal), polycyclic aromatic hydrocarbons (PAHs), a group of carcinogenic chemicals, and carbon monoxide are all found in cigarette smoke, and can adversely impact women's fertility and increase the risk of premature delivery or low birth weight by increasing oxidative stress and inflammation.^13–15 In addition, exposure to volatile organic compounds (VOCs), including those found in cigarette smoke, has been associated with lower birth weight and with a higher rate of birth defects.^16–19 Nicotine, cadmium, and other chemicals present in cigarettes affect the number of follicles in the ovaries and follicular growth.^13,20 Similarly, PAHs block the conversion of androgens to estrogen, which may increase the risk of early menopause.^13,20 The interaction between PAHs and human papillomavirus (HPV) is a key risk factor for HPV-related cancer development.^21–23 Cancer-related risks can be a result of tobacco-specific nitrosamine (TSN) 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), which is converted to the metabolite 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanonol (NNAL).²⁴ Along with nicotine and estrogen, NNK activates transcription factors implicated in the transformation, migration, invasion, and proliferation of malignant cells, which, combined with oxidative stress and individual susceptibility, could explain the higher prevalence of lung cancer in women than in men reported by some studies.^25–28 Cardiovascular risks may be due to oxidizing chemicals, carbon monoxide, heavy metals, and nicotine constituents, which can trigger an inflammatory response and generate sympathetic nervous system activation,²⁹ which leads to cardiovascular disease, the leading cause of death among women.³⁰ Thus, many ingredients in tobacco smoke may cause harm in WRA through endocrine disruption, formation of DNA adducts, oxidative stress and inflammation, and direct toxic effects.

More recently, electronic cigarettes (e-cigarettes) have been introduced as an alternative to conventional cigarettes. Since their arrival at the U.S. market, e-cigarettes have gained significant popularity.^31–33 Approximately 6% of nonpregnant WRA and 5% of pregnant women report using e-cigarettes.^12,34 E-cigarettes are battery-operated devices that heat and aerosolize nicotine in a solution of propylene glycol, glycerin, and flavorings. In contrast to conventional cigarette smoke, the aerosol (commonly referred to as vapor) produced by e-cigarettes is not the product of combustion. This difference, along with fewer chemicals in the e-cigarette solution, suggests that e-cigarette users are exposed to fewer toxicants than cigarette smokers, which has been shown by laboratory studies and U.S. nationally representative surveys.^35,36 However, little is known about the levels of toxicant exposure biomarkers among WRA who use e-cigarettes.³⁶ Nicotine metabolism is higher among WRA than among men and menopausal and postmenopausal women, which could result in modifications to smoking or vaping behavior, and could lead to potentially higher exposure to toxicants.³⁷ Consequently, studies on biomarkers of tobacco-related toxicant exposure can be used to compare exposure from cigarette smoking and e-cigarette use, and to estimate the benefit/risk ratio of these products to one another and to no tobacco use in WRA.³⁸

Moreover, there is a paucity of data in humans and particularly in WRA on whether e-cigarette use is associated with lower levels of inflammation or oxidative stress than conventional cigarette smoking. The existing literature suggests that e-cigarettes affect the immune response with some studies supporting a proinflammatory response,^39–44 while some others have shown evidence of a diminished or impaired response.^45–47 The aim of our study was to characterize biomarkers of toxicant exposure, inflammation, and oxidative stress among WRA who self-report exclusive use of either e-cigarettes or conventional cigarettes; or those who report never using a tobacco product. We hypothesized that WRA who report exclusive e-cigarette use will have higher biomarkers of tobacco-related toxicant exposure than never users. We also hypothesized that exclusive e-cigarette users will have lower levels of biomarkers of tobacco-related toxicant exposure and other biomarkers of inflammation than exclusive conventional cigarette smokers.

Methods

Population

We analyzed restricted biomarker data from the first wave (2013–2014) of the Population Assessment of Tobacco and Health (PATH). PATH is a nationally representative, longitudinal cohort study of tobacco use and health in the United States.⁴⁸ Due to the subsampling of adults at Wave 1 to create the Wave 1 Biomarker Core, PATH biomarker subsample includes a nationally representative of never, current, and recent former (within 12 months) users of tobacco products in the U.S. civilian, noninstitutionalized adult. The initial survey in Wave 1 was conducted between September 12, 2013 and December 14, 2014. Adults were sampled in two phases: Phase 1 sampling used information provided by one adult household member in the household screener, and Phase 2 sampling used information that the sampled adult provided in the Phase 2 screener at the beginning of the adult interview. Computer-assisted personal interviewing was used for the household screener. Audio computer-assisted self-interviewing was used for the interviews. The questionnaire used for Wave 1 can be freely accessed online.⁴⁹ Every respondent who completed an adult interview during Wave 1 was asked to provide biospecimens. Detailed information about the survey has been reported elsewhere.⁴⁸ Of the 32,320 respondents who completed the adult interview, 21,801 (67.4%) provided a urine specimen, and 14,520 (44.9%) provided a blood specimen. PATH researchers selected a sample of 11,522 adults who provided sufficient urine for the planned analyses from a diverse mix of six tobacco product use groups representing never, current, and recent former (within 12 months) users of tobacco products. This group constitutes the Wave 1 Biomarker Core. Of the 11,522 adults, 7,159 also provided a blood specimen. All urine and blood specimens provided by the Wave 1 Biomarker Core were sent for laboratory analysis.^50,51

We included women between the ages of 18 and 49 years, including pregnant women (approximately <4% of the sample) for whom data on biomarker measurements were available. Due to the small number of exclusive e-cigarette pregnant users in the sample, we did not receive authorization to publish the results for that particular group or women.

Exclusive cigarette smokers (“cigarette smokers”) were defined as those who reported smoking >100 cigarettes in their lifetime, currently smoke cigarettes every day or some days, and no current use of e-cigarettes. Exclusive e-cigarette users (“e-cigarette users”) were defined as those who reported using e-cigarettes every day or some days and no current use of conventional cigarettes. Controls were defined as never tobacco users who reported no lifetime use of any tobacco product. We excluded participants who reported “current every day or someday use” or “any use in the last 3 days” of other tobacco or nicotine products, including cigarillos, traditional and filtered cigars, pipe, hookah, smokeless tobacco, dissolvable tobacco, and snus or nicotine replacement therapy. The study sample included 1,857 WRA who had their urine biospecimens analyzed (Fig. 1). This group of women represents a subcohort of the adult population described by Goniewicz et al. previously.³⁶

FIG. 1.

Flowchart of sample selection for analysis of urine toxicants among WRA. WRA, women of reproductive age.

A total of 1,053 of these women also had their blood biospecimens analyzed (Fig. 2). The study was determined to be exempt by the UConn Health IRB.

FIG. 2.

Flowchart of sample selection for analysis of inflammatory and oxidative stress markers among WRA.

Biospecimen processing and laboratory measurements

Urine samples

Full-void urine specimens were self-collected by participants in 500 mL polypropylene containers, which were immediately placed in a custom Crēdo Cube shipper certified to hold contents between 2°C and 8°C for ∼72 hours. Samples were shipped overnight to the PATH biorepository. Each specimen was processed at the biorepository and divided into multiple aliquots totaling to up to 50 mL for each participant. Aliquots were placed in long-term storage at −80°C until further processing. Biomarkers were subsequently measured using highly selective mass spectrometric methods at the Centers for Disease Control and Prevention's (CDC) Division of Laboratory Sciences. Information on laboratory procedures and assessments of urine biomarkers and toxicant levels can be found elsewhere.^50,52 Participants with urine samples were invited to provide blood samples.

Blood samples

Blood was collected from adults at a separate visit by a phlebotomist who visited the respondent's home after the interview. Collected blood was immediately placed in a Crēdo Cube shipper and shipped overnight to the biorepository. All blood aliquots were placed in long-term storage at −80°C until further processing. Information on laboratory procedures and assays can be found elsewhere.⁵⁰ All participants were invited to provide blood samples. Assays for plasma fibrinogen, serum interleukin 6 (IL-6), and serum soluble intercellular adhesion molecule (sICAM) were conducted by a third party (GenWay Biotech, Inc.).^50,52 Serum highly sensitive C-reactive protein (hsCRP) levels were measured by the National Center for Environmental Health (NCEH), CDC.^50,52

Other data

Data were collected on age, race, education, weekly exposure to secondhand cigarette smoke, other tobacco products, and marijuana use. Weekly exposure to secondhand cigarette smoke was defined as the “number of hours in past 7 days that you were in close contact with others when they were smoking.” Marijuana users reported having used “marijuana, hash, THC, grass, pot, or weed” within the past 30 days.

Outcomes

The following urine biomarkers of tobacco toxicant exposure were selected based on their potential harm to women's health (Table 1), including reproductive outcomes: (1) metals: cadmium and lead; (2) PAHs: 1-hydroxypyrene and 2-hydroxynaphthalene (2-naphthol); (3) TSNs: NNAL; (4) nicotine metabolites: cotinine and trans-3′-hydroxycotinine; and (5) VOCs: N-acetyl-S-(2-cyanoethyl)-L-cysteine (acrylonitrile), N-acetyl-S-(2-carboxyethyl)-L-cysteine (acrolein), and N-acetyl-S-(2-carbamoylethyl)-L-cysteine (acrylamide).

Table 1.

Toxicant and Biomarkers of Exposure

Toxicant	Biomarker	Health effects	Reference
Heavy metals	Cadmium	Endocrine disruptor associated with decreased follicles, early menarche, infertility, miscarriage, and pre-eclampsia.	88–94
	Lead	Associated with early menopause, miscarriage, still birth, and congenital defects. Lead can be mobilized from pregnant women into the fetus leading to fetal neurotoxicity.
Polycyclic aromatic hydrocarbons	1-Hydroxypyrene	Associated with decreased fertility and decreased pregnancy viability; and reports of having a role in the development of cervical cancer. and early menopause.	22,23,95
	2-Hydroxynaphthalene (2-naphthol)
Tobacco-specific nitrosamines	NNAL	TSNs are carcinogenic and mutagenic. Lung, nasal, and oral cancer have been associated with DNA adducts produced by TSNs	24
Nicotine	Cotinine	Nicotine is the main addictive substance in cigarettes. Women metabolize nicotine faster than men. Thus, women may be more susceptible to nicotine addiction and have a more difficulty quitting smoking. Nicotine is a neuroteratogen and may play a role in cardiovascular disease.	96–98
	Trans-3′-hydroxycotinine
VOCs	N-Acetyl-S-(2-cyanoethyl)-L-cysteine (acrylonitrile)	Exposure to VOCs has been associated with low birth weight and with a higher rate of birth defects	16–18
	N-Acetyl-S-(2-carboxyethyl)-L-cysteine (acrolein)
	N-Acetyl-S-(2-carbamoylethyl)-L-cysteine (acrylamide)

NNAL, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanonol; TSN, tobacco-specific nitrosamine; VOC, volatile organic compound.

We also selected biomarkers of inflammation and oxidative stress that may be affected by toxicant exposure: (1) urine prostaglandin F2a-8-isoprostane (F2PG2a), (2) plasma fibrinogen, (3) serum hsCRP, (4) serum IL-6, and (5) serum sICAM. Only urine samples from subjects who also provided a blood specimen were analyzed. PATH study designers made this decision to ensure that sample sizes were roughly equivalent to the other blood-based assays, and to ensure that weights adjusting for varying selection probabilities and differential nonresponse rates in blood-based assays could be used to analyze F2PG2a data. Nicotine metabolite ratio (NMR) in the urine was calculated as the ratio of trans-3′-hydroxycotinine to cotinine.⁵³

Statistical analysis

For differences in baseline characteristics by tobacco use group, continuous variables were analyzed using Student's t-test, and categorical variables were analyzed using chi-square test. The biomarkers were log transformed because of skewness of the data. Urine biomarkers were adjusted for urine creatinine to account for differences in hydration status. Multivariable linear regression and Wald's tests were used to determine the differences in the geometric means of the biomarkers between the different tobacco use groups after adjusting for all the confounders in Table 2, except daily product use. Daily product use was only adjusted for in the models that compared exclusive cigarette users with exclusive e-cigarette users. Weighting procedures adjusting for varying selection probabilities and differential nonresponse rates were included in the analysis. Weights differed for urine and blood (plasma and serum) biomarkers. Details about the weights can be found elsewhere.⁵¹ Participants with missing values were removed from the analysis. The analysis was performed using R 3.4.2.⁵⁴ A p-value of <0.05 was considered significant.

Table 2.

Sociodemographic Characteristics of Women of Reproductive Age in Wave 1 of the Population Assessment of Tobacco and Health Study

	Cigarette smokers % (95% CI) n = 961	E-cigarette users % (95% CI) n = 109	Controls % (95% CI) n = 787	p
Age (years)				0.04
18–24	16.4 (12.5–20.4)	16.1 (9.0–23.2)	22.9 (19.8–25.9)
25–34	35.6 (30.8–40.4)	36.5 (25.4–47.7)	27.5 (23–32.1)
35–49	48 (42.6–53.3)	47.4 (34.8–60.0)	49.6 (43.9–55.3)
Race				<0.001
White only	85 (82.3–87.7)	84.4 (78.3–90.5)	70.6 (66.3–74.9)
Black only	8.3 (6.5–10.1)	10.2 (4.8–15.6)	17.9 (13.9–22.0)
Other	6.6 (4.4–8.9)	5.4 (1.7–9.1)a	11.5 (8.6–14.5)
Educational attainment				<0.001
Less than high school or GED	18.5 (15.9–21.1)	10.9 (5.5–16.3)	14.1 (10.5–17.7)
High school graduate	27.5 (21.9–33.1)	20.8 (10.9–30.8)	22.9 (18.6–27.2)
Some college (no degree) or associate's degree	41.6 (37.4–45.8)	50.7 (39–62.4)	32.4 (27.3–37.5)
Bachelor's degree or higher	12.4 (8.2–16.6)	17.5 (9.7–25.4)	30.5 (25.6–35.5)
Poverty level				0.01
Below poverty level (<100% of poverty guideline)	38.8 (34.4–43.1)	27.8 (18.6–37.0)	35 (29.5–40.5)
At or near poverty level (100–199% of poverty guideline)	29.2 (23.8–34.6)	26.8 (17.4–36.2)	21.1 (17–25.2)
At or above twice poverty level (≥200% of poverty guideline)	32 (26.6–37.4)	45.4 (36.2–54.6)	43.9 (38.2–49.5)
Marijuana use within past 30 days	17.2 (12.5–21.8)	12.4 (5.5–19.3)	0.6 (0.1–1.2)a	<0.001
Hours past week secondhand smoke exposure (mean)b	20.6 (17.3–23.9)	6.4 (4–8.9)	2.1 (1.3–2.9)	<0.001
Daily product usec	79.2 (74.7–83.8)	47.9 (36.8–59.1)	N/A	<0.001
Former cigarette smokerd	N/A	70.8 (58.5–83.0)	N/A
Former cigarette smokere	N/A	51.4 (38.5–64.3)	N/A

^aRelative standard error (RSE) >30.

^bNumber of hours in the past 7 days that you were in close contact with others when they were smoking.

^cReported current daily cigarette smoking or daily e-cigarette use.

^dReporting smoking >100 cigarettes in lifetime and currently does not smoke at all.

^eReported smoking >100 cigarettes in lifetime, currently does not smoke at all, and last smoked with past 12 months.

CI, confidence interval.

Results

Background characteristics of WRA included in the analysis

The sociodemographic characteristics of the 1,857 WRA who met our inclusion criteria are presented in Table 2. Among them, 961 were cigarette smokers, 109 were e-cigarette users, and 787 were controls. Control participants were more likely to be 35–49 years old and to have a bachelor's degree, but less likely to self-identify as White, than cigarette smokers and e-cigarette users. E-cigarette users were more likely to have at least some college or associate degree, and were less likely to be below the poverty level than cigarette smokers and controls. Cigarette smokers were more likely to report marijuana use and have more hours of secondhand smoke exposure than e-cigarette users and controls. Cigarette smokers were more likely to report daily use of their chosen product (cigarettes) than e-cigarette users who reported daily use of their chosen product (e-cigarettes) [79.2% (95% confidence interval, CI: 74.7–83.8.) vs. 47.9% (95% CI: 36.8–59.1)]. Data on biomarkers of inflammation and oxidative stress were available for 1,053 WRA, including 536 cigarette smokers, 74 e-cigarette users, and 443 controls.

Urinary toxicant levels in WRA

The geometric levels adjusted by creatinine clearance are presented in Table 3.

Table 3.

Geometric Means of Toxicant Exposure Compounds in Urine and Differences by Group, Adjusted for Confounders

Toxicant (ng/mg of creatinine)	E-cigarette users (n = 109)	Cigarette smokers (n = 961)	Controls (n = 787)	p*
	Mean (95% CI)	Mean (95% CI)	Mean (95% CI)
Cadmium	0.2 (0.1–0.2)	0.2 (0.2–0.2)a	0.1 (0.1–0.1)a	<0.001
Lead	0.4 (0.3–0.4)b	0.4 (0.4–0.4)a	0.3 (0.3–0.3)a,b	<0.001
2-Naphthol	6.0 (5.4–6.7)c	14.5 (13.0–16.0)a,c	5.8 (5.3–6.4)a	<0.001
1-Hydroxypyrene	0.2 (0.1–0.2)c	0.3 (0.3–0.4)a,c	0.1 (0.1–0.2)a	<0.001
NNAL	0.005 (0.004–0.007)b,c	0.2 (0.1–0.2)a,c	0.0009 (0.0008–0.001)a,b	<0.001
Cotinine	91.9 (34.7–243.2)b,c	1,507.6 (1,067.5–2,129.3)a,c	0.4 (0.4–0.5)a,b	<0.001
Trans-3′-hydroxycotinine	181.5 (76.6–430.2)b,c	2,609.5 (1,842.5–3,695.7)a,c	0.7 (0.6–0.8)a,b	<0.001
N-acetyl-S-(2-carboxyethyl)-L-cysteine (acrolein)	98.7 (84.2–115.7)c	235.0 (208.6–264.8)a,c	87.0 (81.1–93.3)b	<0.001
N-acetyl-S-(2-cyanoethyl)-L-cysteine (acrylonitrile)	3.8 (2.8–5.1)b,c	97.1 (76.3–123.7)a,c	1.2 (1.1–1.3)a,b	<0.001
N-acetyl-S-(2-carbamoylethyl)-L-cysteine (acrylamide)	58.8 (51.2–67.6)b,c	135.1 (122.9–148.4)a,c	44.9 (41.8–48.1)a,b	<0.001
NMRd	2.0 (1.6–2.4)	1.7 (1.6–1.9)

^*Wald's test from multivariable regression adjusted for confounders found in Table 2.

^aStatistically significant difference between cigarette smokers and controls (p < 0.001).

^bStatistically significant difference between e-cigarette users and controls (p ≤ 0.01).

^cStatistically significant difference between cigarette smokers and e-cigarette users (p < 0.001).

^dNMR calculated as the ratio of trans-3′-hydroxycotinine to cotinine.

NMR, nicotine metabolite ratio.

E-cigarette users versus controls

The geometric mean for lead was statistically significantly higher in e-cigarette users compared with controls [0.4 (95% CI: 0.3–0.4) ng/mg creatinine vs. 0.3 (95% CI: 0.3–0.3) ng/mg creatinine, respectively]. The VOCs acrylonitrile and acrylamide, but not acrolein, were statistically significantly higher in e-cigarette users compared with controls [3.8 (95% CI: 2.8–5.1) ng/mg creatinine for acrylonitrile, 58.8 (95% CI: 51.2–67.6) ng/mg creatinine for acrylamide, and 98.7 (95% CI: 84.2–115.7) ng/mg creatinine for acrolein vs. 1.2 (95% CI: 1.1–1.3) ng/mg creatinine for acrylonitrile, 44.9 (95% CI: 41.8–48.1) ng/mg creatinine for acrylamide, and 87.0 (95% CI: 81.1–93.3) ng/mg creatinine for acrylonitrile, respectively]. The geometric mean of NNAL was higher among e-cigarette users than controls [0.005 (95% CI: 0.004–0.007) vs. 0.0009 (95% CI: 0.0008–0.001)]. The nicotine metabolites cotinine and trans-3′-hydroxycotinine were also statistically significantly higher in e-cigarette users than in controls [91.9 (95% CI: 34.7–243.2) ng/mg creatinine and 0.4 (95% CI: 0.4–0.5) ng/mg creatinine vs. 181.5 (95% CI: 76.6–430.2) ng/mg creatinine and 0.7 (95% CI: 0.6–0.8) ng/mg creatinine, respectively].

E-cigarette users versus cigarette smokers

The geometric means for cadmium, lead, and NMR were comparable across both groups (Table 3). The geometric means for the PAHs 2-naphthol and 1-hydroxypyrene were statistically significantly higher among cigarette smokers compared with e-cigarette users [14.5 (95% CI: 13.0–16.0) ng/mg creatinine vs. 6.0 (95% CI: 5.4–6.7) ng/mg creatinine, and 0.3 (95% CI: 0.3–0.4) ng/mg creatinine vs. 0.2 (95% CI: 0.1–0.2) ng/mg creatinine, respectively]. Levels of VOCs acrolein, acrylonitrile, and acrylamide were also statistically significantly higher in cigarette smokers than in e-cigarette users [235.0 (95% CI: 208.6–264.8) ng/mg creatinine, 97.1 (95% CI: 76.3–123.7) ng/mg creatinine, and 135.1 (95% CI: 122.9–148.4) ng/mg creatinine vs. 98.7 (95% CI: 84.2–115.7) ng/mg creatinine, 3.8 (95% CI: 2.8–5.1) ng/mg creatinine, and 58.8 (95% CI: 51.2–67.6) ng/mg creatinine, respectively]. NNAL, cotinine, and trans-3′-hydroxycotinine were also statistically significantly higher among cigarette smokers than e-cigarette users (Table 3).

Cigarette smokers versus controls

Cigarette smokers had statistically significantly higher geometric mean levels for all urinary toxicant levels when compared with controls. Cotinine and trans-3′-hydroxycotinine levels were statistically significantly higher among cigarette smokers than controls [1,507.6 (95% CI: 1,067.5–2,129.3) ng/mg creatinine vs. 0.4 (95% CI: 0.4–0.5) ng/mg creatinine and 2,609 (95% CI: 1,842.5–3,695.7) ng/mg creatinine vs. 0.7 (95% CI: 0.6–0.8) ng/mg creatinine, respectively].

Markers of inflammation and oxidative stress by tobacco use status

The geometric levels of hsCRP, IL-6, sICAM, fibrinogen, and F2PG2a are presented in Table 4.

Table 4.

Geometric Means of Inflammatory and Oxidative Stress Biomarkers by Group, Adjusted for Confounders

Parameter	E-cigarette users (n = 74)	Cigarette smokers (n = 536)	Controls (n = 443)	p*
	Mean (95% CI)	Mean (95% CI)	Mean (95% CI)
hsCRP ng/mL	1,312.2 (880.8–1,954.7)	1,879.4 (1,583.2–2,231.1)	1,893.8 (1,585.1–2,262.7)	0.27
IL-6 pg/mL	1.4 (1.2–1.7)	1.7 (1.5–1.8)	1.3 (1.2–1.5)	0.18
sICAM ng/mL	208.6 (191.5–227.3)a	254.2 (240.6–268.6)a,b	209.7 (197.7–222.4)b	0.02
Fibrinogen mg/mL	3.1 (2.9–3.3)	3.3 (3.2–3.4)	3.2 (3.2–3.3)	0.59
F2PG2a (8-isoprostane) ng/mL**	0.5 (0.4–0.5)a	0.6 (0.5–0.6)a,b	0.4 (0.4–0.4)b	<0.001

^*Adjusted for confounders found in Table 2.

^**F2PG2a was measured in urine samples, but sampling and weights were made to match serum weights.

^aStatistically significant difference between e-cigarette users and cigarette smokers (p ≤ 0.03).

^bStatistically significant difference between cigarette smokers and controls (p ≤ 0.04).

F2PG2a, prostaglandin F2a-8-isoprostane; hsCRP, highly sensitive C-reactive protein; IL-6, interleukin 6; sICAM, soluble intercellular adhesion molecule.

E-cigarette users versus controls

There were no significant differences in the geometric means of hsCRP, IL-6, sICAM, fibrinogen, and F2PG2a between e-cigarette users and controls.

E-cigarette users versus cigarette smokers

The geometric means for sICAM and F2PG2a were statistically significantly higher among cigarette smokers compared with e-cigarette users [254.2 (95% CI: 240.6–268.6) ng/mL and 0.6 (95% CI: 0.5–0.6) ng/mL vs. 208.6 (95% CI: 191.5–227.3) ng/mL and 0.5 (95% CI: 0.4–0.5) ng/mL, respectively]. We did not find a statistically significant difference for hsCRP, IL-6, or fibrinogen between these two groups.

Cigarette smokers versus controls

The geometric means for sICAM and F2PG2a were statistically significantly higher for smokers compared with controls [254.2 (95% CI: 240.6–268.6) ng/mL and 0.6 (95% CI: 0.5–0.6) ng/mL vs. 209.7 (95% CI: 197.7–222.4) ng/mL and 0.4 (95% CI: 0.4–0.4) ng/mL, respectively]. We did not find statistically significant differences in the geometric means for hsCRP, IL-6, or fibrinogen between these two groups.

Discussion

To our knowledge, our study is one of the first to report toxicant, inflammation, and oxidative stress biomarkers among WRA who exclusively use e-cigarettes, cigarettes, or no tobacco products. We focused on biomarkers that have been specifically implicated in disease among WRA, and could impact women's health through the pathways of inflammation and oxidative stress. Our analysis showed that among WRA, self-reported cigarette-only smokers had a higher level of exposure to most analyzed toxicants than self-reported e-cigarette-only users and controls. Consistent with this finding, smokers had higher levels of the oxidative stress biomarker (F2PG2a) and one marker of inflammation (sICAM) than both e-cigarette users and controls. E-cigarette users had higher levels of several measures of toxicant exposure (including nicotine metabolites, NNAL, lead, and VOCs) than controls.

These findings suggest that exclusive e-cigarette use may reduce toxicant exposure compared with cigarette smoking, but e-cigarettes result in more toxicant exposure compared with no tobacco use. Although the reduced levels of these toxicants in e-cigarette users as compared with cigarette smokers would suggest that by switching from cigarettes to e-cigarettes there may be a beneficial effect. We recognize that toxicants are still present, and even low levels may lead to potentially harmful health outcomes.

This is particularly relevant for WRA who are never smokers and become e-cigarette users. These women may be exposed to toxicants in e-cigarettes shown to be elevated in our study (lead, VOC, and nicotine and its metabolites), which have been associated with nicotine addiction and adverse pregnancy outcomes and negative effects on the offspring.^55–57 Other adverse effects in WRA are outlined in Table 1. Furthermore, it is likely that there are many more toxicants unaccounted in the PATH samples, and despite subjects reporting being exclusive e-cigarette users, the type of device and the composition of the e-juice largely remain unknown in our cohort. This is relevant since the manufacturing of e-cigarette and e-juice is not currently regulated, reflecting significant discordances between the labels of the product and its content.⁵⁸ This could also explain why WRA reporting exclusive e-cigarette use had higher levels of NNAL compared with controls similarly to other reports in the literature,^35,39,42,59 although one cannot rule out misclassification bias. Notably, we did not find a significant difference in the levels of inflammatory markers or oxidative stress between WRA who use e-cigarettes and controls.

Our findings are consistent with studies showing that toxicants traditionally associated with cigarettes are found at lower levels in e-cigarette aerosol than conventional cigarette smoke.^35,36 Similar to a study by Goniewicz et al.³⁶ that used PATH data, we found that, among WRA, e-cigarette users had higher concentrations of toxicant exposure than controls, and cigarette smokers had significantly higher levels of all toxicant exposures when compared with e-cigarette users and controls. In contrast to Goniewicz et al.³⁶ who reported that cadmium exposure was lower among e-cigarette users than cigarette smokers, we did not find a significant difference between those two groups in WRA. This discrepancy may have been due to our smaller sample size (109 vs. 247 e-cigarette users), and our exclusion of men. Women have been shown to use tobacco products differently than men,⁶⁰ and sex differences in use and metabolism of toxicants and nicotine could contribute to our different results. Our study excluded dual users of cigarettes and e-cigarettes to examine the effects of e-cigarettes without confounding effects of other tobacco products in WRA. This exclusion limits the generalizability of our findings to WRA dual users. Another study that examined toxicant exposure in the general population of dual users did not observe a reduction in toxicant exposure compared with cigarette smokers.³⁶ However, future research could examine toxicant exposure in WRA who are dual users.

Notably, our study is the first to our knowledge to investigate oxidative stress and inflammation, which may have key influences on women's reproductive health, in cigarette smokers compared with e-cigarette users. WRA who reported exclusively smoking cigarettes had higher concentrations of sICAM and F2PG2a compared with those who reported using only e-cigarettes or no tobacco products, and did not differ between e-cigarette-only and never tobacco users. These findings suggest that conventional cigarettes, unlike e-cigarettes, may be associated with higher systemic inflammatory response and higher oxidative stress in WRA, and are consistent with prior reports, showing that cigarette smoking is associated with higher levels of oxidative stress,^25,61,62 and sICAM^63–65 compared with nonsmokers. Collectively, these findings suggest that, by replacing conventional cigarettes with e-cigarettes exclusively, WRA might lower their levels of oxidative stress and inflammation. Clinically, this reduction could be beneficial for WRA as oxidative stress may have a significant role in cervical cancer,⁶⁶ as well as lung cancer²⁵ and vascular disease.⁶⁷ In addition, oxidative stress affects follicullogenesis,^68,69 pregnancy complications, and overall health.^70,71

Our results should be interpreted with caution and need to be replicated in other e-cigarette studies, particularly those with newer models of e-cigarettes. Recent studies have shown that flavorings in e-cigarettes can increase free radical formation, which could lead to oxidative stress.⁷² Acute e-cigarette use leads to an increase in circulating markers of oxidative stress and inflammation (ICAM-1 and reactive oxygen species),⁴⁴ and may increase platelet microparticles indicating vascular reactivity.⁷³ Nicotine by itself has proinflammatory and anti-inflammatory properties, and controversy still exists about its association with health outcomes,^56,74–76 but it is presumed to have some level of risk or harmful effects, including worse pregnancy outcomes.^77,78

The main strength of our article is the use of a data set that is a nationally representative of never, current, and recent former (within 12 months) users of tobacco products of civilian, noninstitutionalized adults in the United States. We also excluded WRA who reported currently using cigarillos, traditional and filtered cigars, pipe, hookah, smokeless tobacco, dissolvable tobacco, and snus or nicotine replacement therapy to avoid the confounding effect of other tobacco use, and adjusted for other potential confounders in the analyses.

Our approach also has several limitations. One limitation was the effect of recall bias by participants on survey results, and the limitations in the PATH data on the quantity of tobacco use or type of e-cigarette device used. We relied on self-report, and so there might have been misclassification bias and carbon monoxide measurements were not available. In addition, the use of e-cigarette devices and the e-juice currently in the U.S. market has significantly changed from the first- and second-generation devices on the market at the time of PATH Wave 1 data collection, which could limit the generalizability of our findings to current third- and fourth-generation devices.⁷⁹ New e-cigarette devices as well as the e-juice composition appear to deliver nicotine more efficiently than older devices,⁸⁰ which could result in greater toxicant exposure as nicotine is highly addictive and metabolized at a higher rate by WRA.^81,82 Some of the e-cigarette users were also former cigarette users, which may affect the concentration of toxicants in the urine and serum for recent cigarette quitters. However, e-cigarette users reported no current everyday or someday use of cigarettes and also no use of cigarettes in the past 3 days before urine and blood collection, which reduces the concentration of the toxicants from cigarette smoking. Other sources of exposure to toxicants or other inflammatory processes were not measured, and therefore not accounted for in the analysis. Another limitation was that we were unable to ascertain either the quantity or the specific type of e-cigarettes used due to these questions not being asked for all e-cigarette users or many responding “other.” Our analysis might have missed other biomarkers of toxicant exposure important for WRA because they were not available in the PATH data; for example, benzopyrene was not available but has been implicated in cervical cancer,⁸³ adverse pregnancy outcomes,⁸⁴ and infertility.⁸⁵ Similarly, there may be other biomarkers of inflammation that are more relevant to e-cigarettes that were not evaluated in this study. Our approach may have missed biomarkers or toxicants that are unique to e-cigarette use due to their distinct chemical makeup, therefore more research is needed before complete conclusions can be drawn on their toxicity. In addition, the data analysis is cross-sectional, which limits our ability to make a causal inference about tobacco use and biomarkers and despite using all the available data on WRA in PATH survey, there were still a relatively small number of WRA who reported using e-cigarettes. Finally, our approach may have also been affected by multiple testing as we tested multiple biomarkers of toxicant exposure and inflammation in the three groups.

In conclusion, WRA who use only e-cigarettes or never use tobacco had lower concentrations of nicotine metabolites, TSNs, some PAHs, some VOCs, sICAM, and F2PG2a than those who exclusively smoked cigarettes. However, WRA who reported exclusively using e-cigarettes had higher concentrations of lead, nicotine metabolites, and some VOCs in their urine compared with those who reported never use of tobacco. E-cigarette users and never users did not differ on the markers of inflammation and oxidative stress that were tested. Given recent reports of e-cigarette/vaping-related acute lung injury after e-cigarette use,^86,87 prospective studies are urgently needed to determine the impact of tobacco exposure biomarkers on disease outcomes in WRA.

Author Disclosure Statement

Dr. C.O. has received free nicotine and placebo inhalers from Pfizer Pharmaceuticals for an NIH-funded smoking cessation study in pregnant women. The remaining authors report no conflict of interest.

Funding Information

The study was supported by R01 CA207491 (Oncken, PI) under a Minority Supplement for Dr. Perez. and NHLBI–PRIDE AIRE.