The effect of vaping on the human lung microbiota

Abstract

Objective:

Vaping’s perception as a safe method of nicotine consumption has contributed to its widespread use among American youth. Research indicates that serious lung disease termed EVALI (E-cigarette or Vaping Product Use-Associated Lung Injury) can develop from vaping. However, the broader consequences on lung health remain less understood.

Methods:

We evaluated the effects of vaping on college students’ lungs using fractional exhaled nitric oxide (FeNO) scores, metal concentrations, and microbiota composition collected over three months.

Results and Discussion:

Mass spectrometry analysis of vape coils revealed that ceramic coils contained higher levels of metals Mg, Al, Cr, Mn, Fe, Ni, As, Sr, Ag, and Ti, while mesh coils had elevated concentrations of Cu, Cd, Pb, and Sn. Corresponding metals were detected in participants’ exhaled breath, with vaping status and coil type significantly influencing heavy metal profiles, and FeNO contributing among vapers. FeNO levels positively correlated with vaping duration, indicating effects on airway inflammation. Analysis of exhaled microbiota showed that temporal variation (sampling month) and, among vapers, duration of vaping had stronger influences on microbial composition than vaping status or coil type, while FeNO had minimal impact. Specific metals, including Al, Fe, Co, Zn, and Zr, were modestly associated with microbial patterns, with Zn and Fe showing the strongest effects.

Conclusions:

These findings identified three interconnected effects of vaping: lung inflammation, heavy metal exposure from heating coils, and changes in lung microbiota. This highlights the need for further research to clarify the mechanisms linking these outcomes.

Introduction

E-cigarette use, otherwise known as vaping, has been increasing among adolescents and young adults in the United States. Current e-cigarette use is reported by 10% of high school students and 18.7% of young adults (Patrick et al., 2025; Birdsey et al., 2023). Early e-cigarette use can be predictive of subsequent traditional cigarette smoking. Individuals who vape more frequently report higher frequency and quantity of cigarette use over time (Doran et al., 2017). Despite documented evidence of harmful effects leading to e-cigarette or vaping product use-associated lung injury (EVALI), college students continue to perceive vaping as harmless to bystanders and safer than conventional cigarettes (Johnston et al., 2003; Hart et al., 2017; Clendennen et al., 2023). However, the long-term effects of vaping on lung health are poorly understood. This study aims to address this gap by investigating the effects of vaping on the lungs and airways, especially regarding inflammation, heavy metal accumulation, and the composition of the lung microbiota.

The human lung microbiota consists primarily of bacteria, with contributions from fungi, viruses, and archaea, and is shaped by continuous exposure to inhaled air and environmental conditions (Chong-Neto et al., 2022; Shukla et al., 2017). Changes in this community are increasingly recognized as relevant to respiratory health and disease. Environmental stressors, including heavy metals, have been shown to alter microbial composition, potentially increasing susceptibility to disease, especially in immunocompromised individuals (Chong-Neto et al., 2022; Gajbhiye et al., 2022; Lima de Silva et al., 2012).

Understanding whether vaping affects the lung microbiota is therefore essential, yet previous studies have not conclusively identified links between vaping and microbial community shifts (Ying et al., 2022).

Airway inflammation is a prominent feature of pulmonary diseases including cystic fibrosis, chronic obstructive pulmonary disease, bronchopulmonary dysplasia, and asthma (Tiddens et al., 2000). Fractional exhaled nitric oxide (FeNO) is a widely used biomarker of airway inflammation, reflecting nitric oxide levels in exhaled breath (13). Elevated FeNO is associated with chronic obstructive pulmonary disease, asthma, and allergic rhinitis (Li et al., 2021; Miskoff et al., 2019). It should be noted that, counter-intuitively, cigarette smoking is associated with reduced FeNO due to suppressed nitric oxide production (McSharry et al., 2005). Thus, FeNO must be interpreted cautiously, but it remains a valuable indicator for assessing whether vaping influences airway inflammatory responses.

A third potential pathway through which vaping may affect lung health is exposure to heavy metals. Conventional cigarette smoking is known to introduce toxic metals into the lungs, contributing to diseases including head and neck cancers (Khlifi and Hamza-Chaffai, 2010; Ashraf, 2012). Vaping devices also contain metal heating components, typically mesh or ceramic coils, that heat e-liquids to generate aerosols (Feeney et al., 2022). Mesh coils commonly include kanthal (FeCrAl) or nichrome (NiCr) alloys, whereas ceramic coils provide higher heat resistance (Omaiye et al., 2021). Previous studies indicated that these heating coils are a likely contributor to the presence of heavy metals in e-cigarette vapor, which expose the lungs to heavy metals when inhaled (Arnold, 2018; Olmedo et al., 2018). The results also showed that the type of heating coil affects the concentrations of metals and other elements in the exhalations.

Although e-cigarette aerosols contain fewer toxic chemicals than the smoke of traditional cigarettes, which includes over 7,000 compounds and at least 70 carcinogens, they still contain potentially harmful substances (Facts about E-Cigarettes, 2023). These include propylene glycol (PG), a food additive; carcinogens like acetaldehyde and formaldehyde; acrolein, a herbicide that can cause lung damage; diacetyl, linked to “popcorn lung” disease; diethylene glycol, a toxic chemical; heavy metals such as nickel, tin, and lead; cadmium, which causes respiratory problems; benzene, a compound found in car exhaust; and nicotine, a highly addictive compound that affects adolescent brain development (What’s in an E-Cigarette?., 2025; Matthews et al., 2023). E-liquids typically include nicotine, flavoring, PG or vegetable glycerin (VG), menthol variants, and sometimes THC, CBD, or synthetic cannabinoids such as K2 or Spice, which pose significant health risks (Centers for Disease Control and Prevention \(U.S.\), 2019). These constituents undergo thermal decomposition during vaping and contribute to airway exposure.

Given the rising popularity of vaping and its potential to influence airway inflammation, heavy-metal deposition, and the lung microbiota, a comprehensive assessment of these interactions is needed. The lung plays a central role in aerosol exposure during vaping, serving both as the site of microbial residence and the entry point for nicotine and other inhaled substances. Despite growing interest in the lung microbiota and its relevance to cancer, immunity, and disease, the mechanisms by which vaping may alter this community remain unknown (Natalini et al., 2023; Casey et al., 2020). This pilot study therefore investigates the interrelationships among vaping behaviors, airway inflammation (FeNO), heavy-metal exposure, and lung microbiota composition to provide foundational insight into how vaping may affect respiratory health.

Materials and methods

Study design and participants

The study was conducted from February to April 2024 in a Biosafety Level 2 (BSL2) Model Hospital Chamber at Texas A&M University (College Station, Texas). The chamber (16.4 m³) provided a controlled environment with 6 air changes per hour (ACH) to decontaminate the air between sampling events (Baig et al., 2022). To account for the shift in temperature and humidity across the study period, a temporal variable Month was defined as February, March, and April.

Undergraduate students from Texas A&M University were recruited through the bulk mail system, flyers, and word of mouth. Eligible participants were required to be university students of legal age for e-cigarette use and classified as either vapers or non-vapers (no use of inhalatory drugs within three months prior to the study). All participants provided electronic informed consent. The Institutional Review Board (IRB) of Texas A&M University reviewed and approved this study (IRB2023–0651D).

A total of 30 participants were enrolled, comprising 15 regular vapers and 15 non-vapers. This sample size was deemed sufficient to generate relevant statistical significance for the purposes of a pilot study (Hertzog, 2008). The average age of participants was 21.83 years. The sample population consisted of 66% female, but the vaping group had only 53% female. The vapes used by participants varied in make and model, with 10 having mesh coils and 5 having ceramic coils. The average reported vaping duration was 2.49 ± 2.54 years.

Experimental setup

Participants were instructed to bring the e-cigarette and e-liquid they routinely used. This led to a variety of device models, flavors, and active ingredients; one participant utilized a cannabidiol (CBD) product, which has no legal age restriction in Texas. Such variety was allowed in the study to address potential configurations involved in vaping. The study focused on differences in coil composition due to its potential influence on heavy metal vapor content (Arnold, 2018).

Air sampling was conducted using the Wetted Wall Cyclone (WWC) collector operated at 100 L/min to preserve cell integrity (King and McFarland, 2012; Zhang et al., 2025). The WWC was first run for 15 min before the participants entered the chamber to record a baseline aerosol profile. The participants then entered the chamber and vaped for 15 min, following their typical frequency and duration of inhalations, while exhaling toward the WWC inlet. After participants exited the chamber, air was collected for additional 15 min post-vaping. Three air samples were collected from each participant: Before (baseline), during and after sample. In the analysis, baseline values were subtracted from subsequent measurements to account for background contamination. Participants were instructed to exhale directly toward the WWC inlet throughout the sampling period, regardless of whether they had taken a recent inhalation from their device. Although individual breathing patterns were not quantitatively measured, this standardized instruction ensured that the WWC captured the totality of each participant’s exhaled breath during the session.

FeNO levels were measured immediately after each vaping event using a standardized protocol with a handheld electrochemical analyzer device (NIOX Group PLC; Oxford, United Kingdom) in compliance with the American Thoracic Guidelines (ATS) (Hanania et al., 2018). As underlying health conditions and individual physiological factors can influence FeNO levels and may contribute to the variability, the test was postponed if the participant had any symptoms of respiratory infection (e.g. a cold or flu). Participants inhaled and exhaled through a filtered aperture, and the device was calibrated before testing. FeNO concentrations below 25 ppb, 25–50 ppb, and over 50 ppb were categorized as normal, intermediate, and high, respectively.

Coil composition analysis

To characterize coil composition in the vaping devices, two devices—a mesh coil (Geek Bar) and a ceramic coil (Vozol)—were deconstructed, due to their frequent (>50%) use within the participants. Coils were sectioned into identical lengths and tested in triplicate. The ceramic coil included a porous ceramic block and a metallic heating element, which were sectioned into six samples—three samples included both components, and three included only the ceramic block. The mesh coils were homogeneous in composition.

Elemental composition was quantified using inductively coupled plasma mass spectrometry (ICP-MS) with a Perkin Elmer Nexion 300D spectrometer (Perkin Elmer, Waltham, MA) following the EPA method 6010 C. WWC sample aliquots (100 μL) were dissolved in nitric acid to prepare a 10 mL solution with a final 1% nitric acid concentration. Coil fragments (0.05 g) were placed in 10 mL of 1% nitric acid overnight. The nitric acid solutions were then analyzed.

Bacterial quantification and identification by qPCR and sequencing

Whole-cell quantitative Polymerase Chain Reaction (qPCR) was used to determine the total bacterial concentration in the aerosol samples as previously described (Smith et al., 2025). The reaction mixture contained the 2× Power SYBR Green PCR Master Mix (Applied Biosystems, Warrington, UK), the bioaerosol sample, and the 16S rRNA primers (1369 forward 5′-AAGTCGTAACAAGGT-3′ and 1492 reverse 5′-ACCTTGTTACGACTT-3′) to amplify a 123 bp fragment (Suzuki et al., 2000). E. coli genomic DNA and nuclease-free water were used as positive and negative controls, respectively. A standard curve generated with E. coli MG1655 was used to convert cycle threshold (Ct) values to genome copy number (GCN), assuming an average of four 16S rRNA copies per bacterial genome (Estrada-Perez et al., 2018). Amplification was performed on the AB StepOne RT-PCR System (AB, Foster City, CA) with an initial cycle of 95 °C for 10 min followed by 40 cycles at 95 C for 15 s and 60 °C for 60 s. A dissociation curve (melting curve) was constructed between 60 °C and 95 °C to ensure amplificon specificity.

Total genomic DNA was extracted using an alkaline lysis method (Zhou et al., 1990). PCR was used to amplify the V4/V5 regions of the 16S rRNA gene (Suzuki et al., 2000). The amplicons were sequenced on the Illumina platform by the Texas A&M Institute for Genome Science and Society (TIGSS). The Bacterial and Viral Bioinformatics Resource Center (BVBRC) open-source software along with ALADDIN Bioinformatics built on the QIIME pipeline were used to delineate the microbiota composition (Caporaso et al., 2010).

Statistical analysis

A 95% significance threshold was used. One-sided independent sample t-tests were conducted using JASP 0.18.2 for analyzing metal concentrations in exhalation samples between vapers and non-vapers. The Mann–Whitney test was used to ensure robustness. Spearman’s rank correlation coefficient was calculated to assess the relationship between the length of time spent vaping and FeNO score (Akoglu, 2018). Principal Coordinate Analysis (PCoA) was performed on the Bray–Curtis distance matrix between metal concentration or microbiota composition and variables, followed by Permutational Multivariate Analysis of Variances (PerMANOVA) to determine statistical significance as previously described (Zhang et al., 2025).

Results

Participant information

Of the vaping participants, 46% reported they had tried to quit using e-cigarettes at least once, with 57% of those reporting they had attempted to quit 3–5 times. Of the vaping participants, 86% reported vaping at least once per day, with 60% reporting vaping many times per day. All relevant basic participant information is presented in Table 1.

Table 1.

Participant information.

Participant #	Age	Gender	Vape	Coil	Duration (year)	FeNO
1	21	Male	Vozol	Ceramic	7	22
2	34	Female	N/A	N/A	0	11
3	29	Female	N/A	N/A	0	17
4	21	Female	N/A	N/A	0	18
5	21	Female	N/A	N/A	0	14
6	21	Female	Lost Mary	Mesh	1.13	7
7	19	Male	Juul	Mesh	3	31
8	18	Female	Vuse	Ceramic	3.73	12
9	30	Female	North	Mesh	1.51	18
10	18	Male	Extract Labs	Ceramic	4.33	28
11	20	Female	Esco	Mesh	1.51	10
12	22	Female	N/A	N/A	0	14
13	20	Female	N/A	N/A	0	27
14	20	Female	N/A	N/A	0	9
15	22	Female	N/A	N/A	0	8
16	21	Female	N/A	N/A	0	14
17	22	Male	Caliburn	Mesh	6.44	16
18	19	Female	Lost Mary	Mesh	1.03	12
19	22	Female	Orion	Mesh	7.32	59
20	22	Female	N/A	N/A	0	12
21	22	Male	Vuze	Ceramic	5.6	33
22	22	Male	N/A	N/A	0	15
23	21	Female	N/A	N/A	0	40
24	22	Female	N/A	N/A	0	6
25	21	Male	Airbar	Mesh	6.46	10
26	20	Female	Juicy Bar	Mesh	0.48	14
27	21	Female	Vozol	Ceramic	1.11	11
28	21	Male	Geek	Mesh	1.06	17
29	22	Male	N/A	N/A	0	48
30	21	Male	N/A	N/A	0	9

FeNO

There was a moderate to strong positive correlation between the duration of vaping and the inflammation of participants’ airways based on the FeNO score. We found Spearman’s ρ to be 0.382 with a p-value of 0.022, while Pearson’s r value was 0.569 with a p-value of <0.001. This indicates that vaping and airway inflammation are linked (Figure 1).

Figure 1.

Graphical representation of the correlation between duration of vaping and FeNO score. Non-vapers were assigned a value of 0.

ICP-MS

Coil

Performing ICP–MS on the deconstructed vapes revealed different elemental compositions. The ceramic coil with heating element displayed high levels of several metals, including magnesium, aluminum, chromium, manganese, iron, nickel, arsenic, strontium, and titanium. The mesh coil presented stronger copper and lead levels, with similar zinc content. This information was used later in the comparative analysis of the heavy metal concentrations in the participants’ lungs (Table 2).

Table 2.

Concentration of metals in ng per mg of coil. Sample background has been subtracted.

Metal	Coil type	Concentration (ng/mg coil)
Zirconium	Mesh	0.016
	Ceramic (with heating element)	0.463
	Ceramic (no element)	0.067
Lead	Mesh	0.721
	Ceramic (with heating element)	0.071
	Ceramic (no element)	0.077
Strontium	Mesh	0.031
	Ceramic (with heating element)	0.047
	Ceramic (no element)	0.122
Arsenic	Mesh	0.005
	Ceramic (with heating element)	0.151
	Ceramic (no element)	0.109
Zinc	Mesh	3.052
	Ceramic (with heating element)	6.867
	Ceramic (no element)	6.940
Copper	Mesh	1.661
	Ceramic (with heating element)	0.267
	Ceramic (no element)	0.328
Nickel	Mesh	0.881
	Ceramic (with heating element)	42.68
	Ceramic (no element)	1.042
Cobalt	Mesh	0.026
	Ceramic (with heating element)	0.141
	Ceramic (no element)	0.073
Iron	Mesh	43.82
	Ceramic (with heating element)	262.7
	Ceramic (no element)	176.5
Chromium	Mesh	1.174
	Ceramic (with heating element)	4.352
	Ceramic (no element)	1.699
Calcium	Mesh	10.39
	Ceramic (with heating element)	103.7
	Ceramic (no element)	48.89
Aluminum	Mesh	0.000
	Ceramic (with heating element)	122.9
	Ceramic (no element)	3.831

Lung exhaled air samples

Of the 30 participants’ exhalation samples, an average of 42.3 (47%) contained detectable levels of the metals identified from the coil by ICP–MS. Of those, zinc (Zn) and lead (Pb) were the first and third most frequently occurring metals (95.5% and 76.6%) and were predictable by status as a vaper or non-vaper. Vapers had statistically significantly higher concentrations of zinc and lead in their exhalations (p = 0.0385 for zinc; p = 0.0147 for lead). The normalized metallic composition of each participant’s exhalation is shown in Figure 2.

Figure 2.

Normalized metallic composition of each participant’s exhalation samples. A total of 60 samples is displayed for the 30 different participants. Each participant has 2 samples associated with them: the sample during the procedure (D), and the sample after vaping (A). Shown here are the during and after (D and A) samples, with the before (B) data subtracted from each to account for environmental effects.

PCoA of exhaled heavy metal concentrations measured was performed based on the Bray–Curtis distance matrix. Because coil type and vaping duration were only applicable to vaping participants, two models were constructed. One PCoA included all participants (n = 30) to assess association by vaping status (vaping vs. non-vaping), FeNO score, and sampling month; variables specific to vapers (coil type and duration) were excluded because they were undefined for non-vapers. The second PCoA was restricted to vapers only (n = 15) and included coil type, duration, FeNO score, and sampling month.

PerMANOVA was used to determine statistical significance as shown in Table 3. In the model with all participants (n = 30), the exhaled heavy metal profiles was found to be statistically significantly different between vaping status (F = 2.83, p = 0.013), while FeNO category did not show significant associations (p = 0.117). Sampling month showed some effects on heavy metal profiles (p = 0.100), although did not pass the significance threshold of 0.05. Consistent with these findings, PCoA plot showed clustering of vapers profile and partially separated from non-vapers along the primary ordination axis (Figure 3(a)).

Table 3.

PerMANOVA Results on heavy metal concentration.

Model	Variable	F	p-value
All participants	Status (Vaper vs. Non-vaper)	2.83	0.013
	Month (Feb, Mar, April)	1.65	0.100
	FeNO (Low, Mid, High)	1.62	0.117
Vapers only	Coil (Mesh vs. Ceramic)	2.67	0.030
	Duration (Low, Mid, High)	1.12	0.342
	Month (Feb, Mar, April)	1.54	0.134
	FeNO (Low, Mid, High)	4.62	0.001

Figure 3.

PCoA of exhaled heavy metal concentrations measured by ICP-MS. (a) Full dataset including all participants (n = 30) with vaping status represented by color, sampling month by shape, and FeNO score by size. (b) Partial dataset including vapers only (n = 15) with coil type by color, sampling month by shape, and FeNO score by size. FeNO score with 0.33 standard deviations below the mean were considered low and values with 0.33 standard deviations above the mean were considered high.

Within the vaper-only model (n = 15), heavy metal profiles differed significantly by coil type (F = 2.67, p = 0.030), while vaping duration and sampling month were not statistically significant predictors. FeNO category, however, was significantly associated with variation in metal profiles (F = 4.62, p = 0.001). The corresponding PCoA plot showed visible clustering by FeNO levels and partially by coil type across the first two axes (Figure 3(b)).

These findings indicate that both vaping status and coil type contribute to differences in exhaled heavy metal composition, with FeNO also playing a notable role among vapers, whereas sampling month and self-reported duration of vaping showed minimal association in this study.

Enrichment ratio

Figure 4 shows the increasing log ratio of GCN to CFU (enrichment ratio) of the aerosol samples collected during and after vaping in correlation to ambient temperature and relative humidity. Positive enrichment ratios indicate reduced culturability. In some cases, the numbers are negative, indicating lower GCN values compared to CFU that may be due to the presence of PCR inhibitors from the environment in the as-collected aerosol samples used in whole cell PCR.

Figure 4.

Temperature and relative humidity prior to sampling shown with the enrichment ratio (log (GCN/CFU)) of each sample. No colonies were detected for samples collected on 2/7.

Microbiota analysis

16S rRNA gene sequencing of the exhalation samples revealed a broad range of bacterial taxa. Because many of these taxa were also present in the surrounding environment, we implemented an environmental background–filtering step to distinguish true exhaled microbiota from co-collected environmental contaminants. All taxa identified in matched environmental control samples were removed from participant datasets, ensuring that the remaining taxa represented bacteria detected exclusively in exhalation samples. The outcomes of this filtering process, including taxa detected before and after background removal for each participant, are presented in Supplementary Figures S1–S15.

Two PCoA were performed on the exhaled microbiota in a similar manner. In the model with all participants (n = 30), PerMANOVA indicated that microbial community composition did not differ significantly by vaping status or FeNO levels, whereas sampling month was strongly associated with community variation (F = 2.77, p = 0.001; Table 4). In agreement with these results, the PCoA plot demonstrated noticeable clustering by sampling month, with no clear separation between vapers and non-vapers (Figure 5(a)).

Table 4.

PerMANOVA Results on exhaled microbiota.

Model	Variable	F	p-value
All participants	Status (Vaper vs. Non-vaper)	0.89	0.591
	Month (Feb, Mar, April)	2.77	0.001
	FeNO (Low, Mid, High)	1.08	0.297
Vapers only	Coil (Mesh vs. Ceramic)	1.18	0.206
	Duration (Low, Mid, High)	1.71	0.003
	Month (Feb, Mar, April)	2.55	0.001
	FeNO (Low, Mid, High)	1.20	0.176

Figure 5.

PCoA of exhaled microbiota. (a) Full dataset including all participants (n = 30) with vaping status represented by shape, sampling month by color, and FeNO score by size. (b) Partial dataset including vapers only (n = 15) with vaping duration by shape, sampling month by color, and FeNO score by size.

Within the vaper-only model (n = 15), microbial composition differed significantly by vaping duration (F = 1.71, p = 0.003) and by sampling month (F= 2.55, p = 0.001), while coil type and FeNO level did not show significant associations (Table 4). PCoA visualization was consistent with these findings, showing gradients corresponding to sampling month and vaping duration but little evidence of FeNO (Figure 5(b)).

Together, these analyses suggest that temporal variation (sampling month) and, among vapers, duration of vaping exert stronger influences on exhaled microbiota composition than vaping status or coil type, with FeNO showing minimal association in both models.

Exhaled heavy metal concentrations quantified by ICP–MS were evaluated for potential associations with the corresponding exhaled microbiota profiles (Supplementary Figure S16). PerMANOVA results are summarized in Table 5. Several metals, including iron, zinc, and zirconium showed the strongest associations with microbial community composition (F = 1.74–1.82), with statistically significant effects (p = 0.002 for Zn, p = 0.002 for Fe, and p = 0.006 for Zr). However, these effects remained weaker than that of sampling month (F = 2.77), indicating that temporal variation exerted a more substantial influence on microbiota structure than individual metal concentrations.

Table 5.

PerMANOVA results on exhaled microbiota based on exhaled heavy metal concentrations.

	Zr	Pb	Sr	As	Zn	Cu	Ni	Co	Fe	Cr	Ca	Al
F	1.820	1.071	1.101	1.190	1.742	1.081	0.889	1.616	1.807	1.302	1.408	1.536
p	0.006	0.323	0.268	0.149	0.002	0.276	0.679	0.013	0.002	0.077	0.039	0.020

Discussion

The sampling campaign for this study was conducted over a three-month period from February to April 2024, and was followed by analyses examining airway inflammation, exhaled heavy metal concentrations, and exhaled microbiota composition.

A positive association was observed between vaping duration and FeNO levels (Figure 1), suggesting that chronic vape use may influence inflammatory pathways in the respiratory tract. Although elevated FeNO does not directly establish airway injury, the relationship warrants further investigation, especially given that cigarette use has a negative effect on FeNO levels and inflammation (as represented by FeNO levels) is associated with various pulmonary diseases (McSharry et al., 2005). Airway inflammation is linked to remodeling or thickening of the bronchial walls (Tiddens et al., 2000). Whether similar structural or temporal changes occur in the airways of vapers remains unknown and should be explored in future studies.

The analysis of the heavy metal concentrations revealed that multiple metals including aluminum, iron, and copper were present in substantial amounts within the mesh and ceramic heating elements. Lead, a toxic metal, was the sixth most prominent metal in the mesh coil. Manganese, zinc, and chromium, which are toxic when inhaled, were also detected in the ceramic coil heating elements (Olmedo et al., 2018). These findings are consistent with previous work demonstrating that metals present in heating elements can be transferred into aerosolized emissions and subsequently inhaled by users (Arnold, 2018). Lead has known effects on immune cell activity, renal function, reproductive function, and neurotransmitter release (Wani et al., 2015). Aluminum, magnesium, zinc, chromium, manganese, cadmium, and arsenic are some examples of elements that are both toxic to inhale and present in the heating coils analyzed in this study (Olmedo et al., 2018; Alasfar and Isaifan, 2021; Guertin et al., 2004; A.f.T.S.a.D. Registry, 2014; Occupational Safety and Health Administration, 2014). The metal content documented here therefore underscores a plausible exposure pathway for users and highlights a public health concern linked to long-term vaping. Future studies could explore whether modifications of the coil materials could impact metal particle transmission. Strengthening regulations or providing industry incentives for safer designs could help mitigate consumer exposure to harmful substances (E-cigarettes: Facts, stats and regulations, 2021).

PCoA of exhaled heavy metal concentrations showed distinct clustering associated between vapers and non-vapers, and among vapers, strong associations were found with FeNO levels and type of coils. This indicated that individuals with middle-to-high FeNO exhibited more similar metal profiles than those with low FeNO. Given that FeNO increased with vaping duration, this pattern suggests that prolonged exposure to vape aerosols may contribute to the accumulation or altered distribution of inhaled metals in the airways.

Exhaled microbiota composition was shown to be influenced most significantly by sampling month for all participants. In contrast, in the vaper-only cohort, vaping duration and sampling month were significant predictors of microbial differences, while coil type and FeNO levels were not.

These findings suggest that environmental or temporal factors contributed to the overall structure of the microbiota, whereas specific vaping behaviors influenced microbial variation within active users. Further analysis revealed significant associations between exhaled microbiota composition and specific metal concentrations, most notably zinc, iron, and zirconium. These metals are known to influence bacterial growth, stress responses, and community structure, explaining the biological plausibility of these associations (Lima de Silva et al., 2012).

Tables S1 and S2 show the top taxa based on relative abundance identified in the lung microbiota of non-vapers and vapers associated with pathogenicity and the potential degradation of vape components (47). Several bacteria in the lung microbiota are pathogenic, causing lung infections (Acinetobacter lwoffii, Corynebacerium aurimucosum, C. tuberculostearicum, E. coli, Nevskia aquatilis or N. soli), associated with non-tuberculous mycobacteria (NTM) (microbacteriaceae), EVALI (Streptococcus) or lung cancer (Bradyrhizobium, Gemmata). Many components in the microbiome are able to mobilize and metabolize heavy metals (Algoriphagus terrigena, Arcticibacter, Craurococcus, Ferruginibacter, Hymenobacter, Pedomicrobium, Turicibacter sanguinis, Yonghaparkia alkaliphile) or fix nitrogen from high N compounds like nicotine, as vaping reduces nitric oxide NO production in lungs (Brevundimonas nasdae). Other bacteria are cellulolytic, degrading the wicking material in atomizers or the microcrystalline cellulose filler in nicotine pouches (Ferrovibrio, Kallotenuales, Rudaea cellulosilytica). Some of the bacteria are harmless mucous membrane or skin commensals (Micrococcus luteus, Moraxella osloensis, Kingella denitrificans). Several species that can degrade hydrocarbons (Sediminibacterium salmoneum) or metabolize nicotine (Sphingomonas melonis) were found in the microbiota of vapers. Interestingly, Devosia, which can degrade the mycotoxin deoxynivalenol inhaled from cannabis products, was detected in a non-vaper lung microbiota. In vapers, higher relative abundance was associated with bacteria that cause respiratory infections or are involved in nitrogen fixation and hydrocarbon degradation. Further investigation is required to determine whether the microbiota changes observed in this study contribute to the vaping-related metabolic reactions or health outcomes.

The first two PCoA axes explained 59% of the variance in heavy metal concentrations, indicating a relatively low-dimensional structure with strong covariation among metals. In contrast, the first two axes explained only 25% of variance in the exhaled microbiota, consistent with the inherently high dimensionality and interpersonal variability characteristic of microbial community data. Thus, the lower percentage for the microbiota reflects the greater complexity and dispersion of microbial profiles, rather than a lack of association with the predictors tested. Although common with airway and environmental microbiome datasets, this suggests that variation might come from factors that are not explained by the predictors included in this pilot study, such as diet, oral microbiome, underlying health status, breathing pattern, or interpersonal microbiome differences. Because baseline FeNO levels were not normalized across participants, differences in underlying airway inflammation (e.g. due to allergies or asthma) may also have contributed to variability in both FeNO and microbiota outcomes. Future studies should incorporate baseline physiological measurements and additional lifestyle or clinical metadata to better account for these potential confounders.

Together, these findings suggest interconnected relationships among vaping behavior, airway inflammation, heavy metal exposure, and exhaled microbiota composition. The potential influence of vaping on the lung microbial community may have important implications for respiratory and systemic health, given the role of the microbiota in lung function and immune response. Further longitudinal studies are needed to determine whether the microbiota changes observed here contribute to vaping-related metabolic reactions, susceptibility to infection, or other health outcomes.

Conclusions

This study demonstrates that vaping is associated with measurable changes in airway inflammation, exhaled heavy metal concentrations, and lung microbiota composition. Our findings suggested that exhaled heavy metal composition was affected by vaping status and coil type, while among vapers duration of vaping exert stronger influences on exhaled microbiota composition. FeNO levels showed notable effects on metals but minimal impact on microbiota. Longer vaping histories were linked to elevated FeNO levels, and metals such as iron, zinc, and zirconium were associated with alterations in microbial communities. While these findings highlight potential respiratory health impacts, the underlying mechanisms remain unclear. Future research should include a larger cohort to investigate causal relationships among vaping behaviors, metal exposure, and microbiota alterations, and explore strategies to reduce harmful metal inhalation from device components.

Data availability statement

The authors will share the data associated with this study upon reasonable request.