Blood Phosphatidylethanol Concentrations Following Regular Exposure to an Alcohol-Based Mouthwash

Abstract

Direct biomarkers of ethanol are used to monitor individuals who are required to abstain from ethanol consumption. In recent years, blood phosphatidylethanol (PEth) has gained acceptance in clinical and forensic contexts as an abstinence marker. Its elimination half-life of several days provides a window of detection of days to weeks. However, there is no research addressing the extent of PEth formation related to extraneous ethanol exposures. To assess the degree of ethanol absorption and subsequent formation of blood PEth related a common extraneous exposure, regular use of an ethanol-containing mouthwash, we recruited 16 participants to gargle with an alcohol-based mouthwash (21.6% ethanol) 4 times daily, for 12 consecutive days. Blood was analyzed for PEth 16:0/18:1 by liquid chromatography–tandem mass spectrometry. Our hypothesis that blood PEth concentrations would not equal or exceed 20 ng/mL was confirmed. Although the data suggest that regular use of mouthwash is unlikely to result in suprathreshold PEth concentrations, this work highlights the importance of considering extraneous ethanol exposures in clinical decision-making and in future research.

Introduction

Certain populations—including healthcare professionals with substance use disorders who are monitored by statutorily mandated professionals’ health programs—are required to abstain from ethanol consumption. These programs typically monitor abstinence through the random and for-cause measurement of direct ethanol biomarkers, most commonly urinary ethyl glucuronide (EtG) and ethyl sulfate (EtS), hair EtG and, more recently, blood phosphatidylethanol (PEth).

PEth is an abnormal phospholipid, formed exclusively in the presence of ethanol. Under ordinary conditions—in the absence of ethanol—phospholipase D (PLD) in erythrocyte cell membranes catalyzes the hydrolysis of phosphatidylcholine to phosphatidic acid and choline. Compared with water, however, ethanol is a preferred substrate for PLD and in its presence phosphatidylcholine is converted to PEth and choline (1).

PEth is not a single molecule; it comprises a group of at least 48 glycerophospholipid homologs, each with a phosphoethanol head group and a unique pair of carboxylic acid side chains. The homologs are labeled according to the number of carbon atoms and double bonds in each carboxylic acid side chain. Thus, for example, PEth 16:0/18:1 denotes the presence of 16 carbon atoms and 0 double bonds in chain 1 and 18 carbon atoms and 1 double bond in chain 2 (2). PEth 16:0/18:1 is the most prevalent homolog in humans and the most common analytic target in commercial testing in the USA (3, 4).

The formation of PEth in the presence of ethanol was first described decades ago (5). Early analytic techniques (e.g., nonaqueous capillary electrophoresis and high-pressure liquid chromatography coupled with evaporative light-scattering detection) were relatively insensitive, with limits of detection (LODs) in the hundreds of nanograms per milliliter for the sum of all PEth homologs (6, 7). This translated to ethanol consumption thresholds of ∼50 g/day over 2 to 3 weeks and windows of detection of ∼2 weeks, making it suitable for detection of sustained heavy drinking or for distinguishing heavy from moderate drinking (8, 9).

The introduction of mass spectrometric-based methods in recent years has allowed for the identification and quantification of individual PEth homologues, with LOD and limits of quantitation (LOQs) in the low single digit nanograms per milliliter. This has translated to the detection of single drinking episodes. For example, in participants who consumed a single dose of ethanol to a target blood alcohol concentration (BAC) of 0.10 g/dL, PEth 16:0/18:1 was detectable for a mean of 9.3 ± 3.0 days (range: 3–12 days) at an LOD of 10 ng/mL (10).

The sensitivity of mass spectrometric methods has led to suggestions that PEth may be an acceptable abstinence marker (2, 10–14). Establishing an ideal abstinence cutoff concentration—one that optimizes sensitivity and specificity concerns—remains unresolved. The 4 accredited US laboratories that perform PEth testing have adopted, by consensus, a common cutoff of 20 ng/mL for clinical and forensic testing to distinguish beverage alcohol consumption from abstinence or incidental exposures.

There is a limited literature suggesting that the 20 ng/mL cutoff is a conservative one. For example, Nalesso et al. (2011) examined PEth 16:0/18:1 in 10 “teetotalers,” using a highly sensitive liquid chromatography high-resolution mass spectrometry method (15). All participants produced negative results at an LOQ of 0.7 ng/mL. Piano et al. (2015) examined PEth 16:0/18:1 in 22 abstainers, defined as having consumed no more than one alcoholic beverage in the previous 2 to 3 years. All participants produced negative results at an LOQ of 8 ng/mL (16). Schrock et al. (2017) examined PEth 16:0/18:1 in 12 abstainers, defined as having consumed no ethanol in the previous 2 weeks. All participants produced negative results at an LOD of 10 ng/mL (10). Nguyen et al. (2018) reported on PEth 16:0/18:1 in four abstainers, defined as having consumed no ethanol in the previous month. All subjects produced negative results at an LOQ of 14.8 ng/mL (17). Kechagias et al. (2015) prospectively evaluated 23 self-identified “social drinkers” before and after 3 months of requested abstinence. At the end of the study period, 18 participants produced PEth concentrations below the LOQ of 3.5 ng/mL. Four additional participants produced PEth concentrations—5.6, 9.8, 18 and 18 ng/mL—that were below 20 ng/mL. One participant produced a PEth of 25 ng/mL (13). An important limitation, however, was the investigators’ inability to exclude surreptitious ethanol consumption.

Concentrations of the PEth species 16:0/18:1 (hereafter referred to simply as PEth) in excess of 20 ng/mL are regarded as prima facie evidence of beverage ethanol consumption (13, 16, 18). However, we were unable to identify any literature on the effects of incidental alcohol exposures on the formation of blood PEth.

Prior to conducting this project, we conducted a literature search for PEth outcomes related to incidental ethanol exposures. Search terms included “phosphatidylethanol OR PEth” in combination with each of the following: “mouthwash,” “incidental,” “environmental” and “extraneous.” Results revealed no studies informing the current question.

Two studies have demonstrated the formation of EtG, EtS or both following the use of ethanol-containing mouthwash (19, 20). The present study was designed to investigate the effects of intensive use of high ethanol content mouthwash on the formation of PEth. Our hypothesis was that four-times-daily use of an ethanol-containing mouthwash (21.6% ethanol) would fail to produce blood PEth concentrations that exceed 20 ng/mL, a commonly used cutoff concentration and presumptive evidence of beverage ethanol consumption.

Methods

Participants

Inclusion criteria included ages 18 to 75 years, inclusive and willingness to abstain from ethanol and nonbeverage ethanol-containing products for the 5 days prior to, and the 12 to 13 days of, the study.

Excluded were individuals with current pregnancy (as determined by point-of-care urinary β-hCG testing) or breastfeeding, those who endorsed at-risk ethanol use (as determined by AUDIT-C) or symptoms of urinary tract infection, and those who endorsed histories of alcohol use disorder, hepatic or renal dysfunction or diabetes mellitus.

All participants provided written informed consent. The study protocol (IRB201901085) was approved by the University of Florida Institutional Review Board. Participants comprised nine males and seven females. Participants were compensated for their involvement in the study.

Study design and procedure

Power analyses indicated that data from 15 individuals would provide sufficient power to test the study hypothesis.

Participants were requested to abstain from ethanol and to refrain from using common ethanol-containing products for the 5 days prior to, and the 12 to 13 days of, the study period. Participants were offered flexible windows for attendance of their second (DAY 6 or 7) and final (DAY 12 or 13) study site visits. They were provided a list of products to avoid, including categories that commonly contain ethanol (Figure 1).

Figure 1.

List of products participants were instructed to avoid.

A schematic of the specimen collection schedule is presented in Figure 2. Participants presented to the study site following a minimum of 5 days of abstinence from ethanol and common ethanol-containing products (DAY 0). Participants provided an unobserved urine sample for analysis of baseline EtG and EtS concentrations. Participants also provided a capillary blood sample, collected onto a dried blood spot (DBS) collection card, for analysis of baseline PEth concentration. Biospecimens were marked with the participants’ identification number and date of collection and stored in a research freezer and at room temperature, respectively. Participants were supplied at DAY 0 with 7 days of supplies: 28 (20 mL) single-use, prefilled bottles of Listerine^® Cool Mint^® (21.6% ethanol), seven (100 mL) urine collection cups (each premarked with the participant’s identification number), a Sharpie marker to date the cups, seven zip locking specimen lab transport bags, an insulated foam container and two cold packs. At the DAY 6/7 study site visit, participants were resupplied with sufficient single-use bottles of Listerine^®, urine collection cups and transport bags to complete the study.

Figure 2.

Specimen collection timeline.

On all study days that participants were not scheduled for a study site visit, they collected a first morning void urine specimen for subsequent analysis of EtG and EtS; marked the collection cup with the date; and placed the cup, in a plastic specimen bag, in their freezer. On all study days, participants gargled with 20 mL of mouthwash for 30 seconds after breakfast, lunch, dinner and at bedtime. Participants were instructed to remedy missed postprandial gargles with subsequent gargles in order to total four gargle episodes per day. Following each gargle, participants were instructed to expectorate their oral contents and refrain from rinsing their mouths for at least one minute. On DAYS 6/7 and 12/13, participants presented to the study center with their urine specimens and all mouthwash bottles. DAY 6/7 and 12/13 urine and capillary blood collections were repeated consistent with DAY 0 procedures.

Verification of adherence to the study protocol

Urinary EtG and EtS analyses (see below for analytic methods) were employed to retrospectively evaluate participants’ attestations of abstinence and thus the integrity of the blood PEth data. Based on our previously published research on ethanol-containing mouthwash and the production of urinary ethanol conjugates (19), we established a posteriori that EtG concentrations >250 ng/mL or EtS concentrations >125 ng/mL would be used as critical cutoffs. Urine conjugate concentrations exceeding these concentrations and blood PEth concentrations exceeding 20 ng/mL would result in interviews with participants about possible nondisclosed episodes of beverage ethanol consumption or extraneous ethanol exposures.

Analyses of EtG, EtS and PEth

Urine EtG/EtS and DBS PEth were analyzed at United States Drug Testing Laboratories (USDTL, Des Plaines, IL). EtG and EtS were analyzed using a method based on previously published reports (21, 22). Briefly, 50 µL of urine was diluted with 1,000 µL of an aqueous internal standard solution that contained 50 ng/mL of EtG-d₅ and ETS-d₅. Separation was accomplished by injecting 7 μL of the prepared specimen on a Phenomenex Hypercarb™ porous graphitic column (50 mm × 2.1 mm × 3 µm particle size) using 0.1% formic acid with 6% acetonitrile as mobile phase A and acetonitrile with 0.1% formic acid as mobile phase B on an Agilent 1200 HPLC system (Wilmington, DE) while holding the column compartment at 60°C. The solvent method was isocratic with 98% mobile phase A and 2% mobile phase B with the flow rate set to 450 µL/min. Detection was accomplished using a AB Sciex 3200 tandem mass spectrometer (Foster city, CA) in the negative electrospray ionization mode (ion spray voltage: −3500V, declustering potential: −35V, collision energy (CE): −26V, and collision cell exit potential (CXP): −26V) monitoring the mass transitions m/z 221.1 → 75.1 and m/z 221.1 → 85.1 for EtG and m/z 226.1 → 75.0 for EtG-d₅. EtS was monitored using the m/z 125.0 → 97.0 and m/z 125.0 → 80.0 mass transitions while EtS-d₅ 130.0 → 80.0 mass transitions. A single point calibration was used (EtG: 500 mg/mL; EtS: 125 ng/mL) with each batch controlled with a set of negative, low, medium and high controls (EtG: 0, 100, 625 and 10,000 ng/mL; EtS: 0, 25, 150 and 10,000 ng/mL). The LODs for EtG and EtS were 50 and 12.5 ng/mL, respectively; LOQs were 100 ng/mL (coefficient of variation (CV) = 0.9%) and 25 ng/mL (CV = 3.2%), respectively; the upper limits of linearity were 10,000 ng/mL for both EtG and EtS; cutoff scores of 100 and 25 ng/mL, respectively, were applied to determine positivity.

PEth specimens were analyzed using a previously published method (23). DBS, rather than venous blood, was collected and analyzed because of its stability at room temperature and its resistance to postcollection synthesis of PEth if exposed to ethanol (23–25). Three standard blood spot punches (3.1 mm) were prepared for each dried human blood spot specimen, calibrator, and control. Each specimen was fortified with the addition of the isotopically labeled internal standard, PEth-d₃₁. The punches were extracted with methanol (1 mL), evaporated under a stream of nitrogen, and the residues were reconstituted in 1.0 mL of mobile phase A (50% 2-mM ammonium acetate: 25% acetonitrile: 25% isopropanol). Mobile phase B was 60:40 acetonitrile/2-propanol. Separation was achieved by injecting 2 μL of the extract on an Eksigent HALO (0.5 mm × 50 mm, 2.7 µm of particle size; Sciex, Foster City, CA) C-8 column held at 30°C using an Eksigent Ekspert MicroLC 200 system (Sciex, Foster City, CA) using a solvent program starting at 30% mobile phase B which was increased to 98% at 1.5 minutes, held at 98% until 3.0 minutes, and decreased to 30% at 3.1 minutes. The detector was a Sciex 6500 tandem mass spectrometer (LC–MS-MS; Foster City, CA) fitted with a SelexION^™ Differential Mobility Selection Device (Sciex; Foster City, CA) in negative electrospray ionization mode (Ion spray voltage: −4,300V and declustering potential: −135V,). The method monitored a single isoform of PEth (palmitoyl/oleoyl), which is the most prevalent PEth species monitoring the m/z 701.5 → 281.3 (CE: −42 and CXP: −21) and m/z 701.5 → 255.3 (CE: −48 and CXP: −7) mass transitions for PEth and m/z 706.5 → 281.3 (CE: −46 and CXP: −23) and m/z 706.5 → 255.2 (CE: −50 and CXP: −19) mass transitions for PEth-d₅. A single point calibrator at 20 ng/mL was used with controls at 0, 8, 25 and 100 ng/mL. The LOD was 1.6 ng/mL, the LOQ was 3.2 ng/mL (CV = 12.2%), and the assay was linear up to 200 ng/mL. A cutoff of 20 ng/mL (CV = 5.1%) was applied to determine positivity.

Statistical analysis

All participants in the sample contributed biological samples for each measurement period, resulting in a complete dataset with no missing values. PEth concentrations that were detectable but below the LOQ were adjusted to their maximum possible value of 3.2 ng/mL. In the face of unknown concentrations, this adjustment provided the most conservative estimate for purposes of comparison to a given critical concentration (i.e., 20 ng/mL).

In addition to concentrations collected at each time point, difference scores were computed to quantify change from baseline to DAY 6/7 and DAY 12/13.

PEth concentrations were analyzed with descriptive statistics, confidence intervals and one sample t-tests. In addition to describing concentrations in terms of standard 95% confidence intervals, we utilized a more stringent 99.9% interval, given the potential implications of clinical decisions that may be informed by these data. One sample t-tests (one-tailed) for each time point (Baseline, DAY 6/7, Day 12/13) compared observed PEth concentrations with a 20 ng/mL constant. Due to the frequency of undetectable concentrations, the data were not normally distributed; thus, log transformation was conducted prior to parametric analysis. ± Difference scores were analyzed with one sample t-tests (two-tailed) comparing change from baseline with 0 ng/mL (i.e., zero change).

Results

Sixteen participants were enrolled in the study. Data from one individual was excluded based on highly positive baseline analyte levels. The 15 remaining individuals completed the full study.

Quantitative PEth analysis

Summarized results, including individual time points and derived measures of change from baseline, are presented in Table I.

Table I.

Mean PEth Concentrations and Changes over Time

Time/Measure	Mean concentration (ng/mL)	SD	95% CI [lower, upper]	99.9% CI [lower, upper]	t (P)
Baseline	3.76	5.54	[1.52, 4.98]€	[0.88, 8.65]€	7.35 (P < 0.0001)¥
DAY 6/7	2.69	3.87	[1.35, 4.00]€	[0.81, 6.64]€	8.68 (P < 0.0001)¥
DAY 12/13	3.03	4.84	[1.23, 4.04]€	[0.71, 7.04]€	8.07 (P < 0.0001)¥
Baseline to DAY 6/7	−1.07	2.62	[−2.52, 0.38]	[−3.08, 0.94]	0.64 (P = 0.5342)∆
Baseline to DAY 12/13	−0.73	3.57	[−2.82, 1.36]	[−3.63, 2.17]	1.00 (P = 0.3350)∆

^€Depict estimates following log transformation to correct skewness.

^¥H₀ = 20 ng/mL.

^∆H₀ = 0 ng/mL.

Across all time points, both standard (95%) and conservative (99.9%) confidence intervals suggested that under study conditions, PEth concentrations are unlikely to exceed 10 ng/mL. Similarly, when means from each time point were compared with a value commonly accepted to indicate recent ethanol consumption (20 ng/mL), results suggested a high probability that PEth concentrations observed under study conditions would be unlikely to exceed this cutoff (all Ps < 0.0001). Change from baseline was similarly assessed. For each measure of change from baseline, both confidence intervals overlapped zero, suggesting little change occurred within either 6/7 or 12/13 days of initiating mouthwash use. Thus, it was unsurprising that the magnitude of both change measures failed to differ significantly from 0 ng/mL (all Ps > 0.137).

Descriptive PEth analysis

As illustrated in Table II, of the 15 participants included in the data analysis, 12 produced PEth concentrations ≤5 ng/mL at all time points, the majority of which were below the LOQ. While all results from the full sample remained below the 20 ng/mL threshold throughout the study period, three participants produced concentrations ≥10 ng/mL. Two individuals produced baseline concentrations ≥15 ng/mL; in both cases, concentrations declined at DAYS 6/7 and 12/13. In the remaining case, a 12 ng/mL concentration was observed on DAY 12 despite relatively low concentrations at baseline and DAY 6 (≤5 ng/mL).

Table II.

Raw PEth Concentrations by Participant

Participant #	Sex	PEth concentration (ng/mL)
		Day 0	Day 6 ± 1	Day 12 ± 1
1	M	ND	ND	ND
2	M	4	<LOQ	ND
3	F	5	ND	ND
4	M	ND	ND	ND
5	M	ND	4	<LOQ
6	F	4	4	12
7	F	<LOQ	<LOQ	<LOQ
8	F	4	4	4
9	F	ND	ND	ND
11	M	ND	ND	ND
12	F	ND	ND	ND
13	M	18	13	14
14	M	<LOQ	ND	ND
15	F	ND	ND	ND
16	M	15	9	8

ND, not detected.

Comparison with urinary conjugates and self-report

EtG/EtS concentrations were consistent with PEth concentrations and self-reports. Maximum EtG and EtG₁₀₀ (EtG adjusted to urinary creatinine of 100 mg/dL) concentrations were 206 and 112 ng/mL, respectively. Maximum EtS and EtS₁₀₀ concentrations were 105 and 94 ng/mL, respectively. All EtG and EtS concentrations were below our a posteriori assumptions that concentrations below 250 and 125 ng/mL, respectively, supported participants’ attestations of abstinence.

In the participant who produced a PEth concentration of 12 ng/mL on DAY 12, urinary EtG and EtS concentrations were below their respective LODs at all time points.

Discussion

There is presently no published literature on the effect of extraneous ethanol exposures on the formation of blood PEth. The aim of this pilot study was to determine the effect of one type of extraneous exposure—four-times-daily use of a high ethanol content mouthwash—on the formation of blood PEth in individuals who abstained from beverage ethanol and other common extraneous ethanol exposures. We hypothesized that regular use of an ethanol-containing mouthwash would fail to result in blood PEth concentrations exceeding the 20-ng/mL threshold accepted as evidence of beverage alcohol consumption. Our results support this hypothesis. Indeed, our 99.9% confidence interval suggests that this pattern of mouthwash use is unlikely to produce PEth concentrations exceeding 10 ng/mL.

Of the 15 participants included, all began the study with PEth concentrations below 20 ng/mL. Of the two participants who began the study with higher-than-expected PEth concentrations, both yielded lower concentrations at subsequent time points, supporting our overall hypothesis. While all participants remained below 20 ng/mL across all study visits, one atypical concentration trajectory bears discussion. One participant produced an elevated PEth concentration on DAY 13 (12 ng/mL), relative to DAYS 0 and 6 (both 4 ng/mL). This increase was seemingly at variance with the participant’s self-reports of no beverage ethanol or restricted product consumption and her daily urinary EtG/EtS concentrations, both of which were below their respective LODs at all time points. There are several potential explanations for this incongruity. The first explanation, analytical error, was eliminated by retesting a second aliquot of DAY 13 blood spot specimen, which reconfirmed at 13 ng/mL. Second, it would have been possible for her to have consumed one or two alcoholic beverages on one or more occasions between DAYS 6 and 13 and still produce negative next-day EtG and EtS results (26). Third, knowingly or unknowingly, she could have consumed nonbeverage sources of ethanol. Fourth, she could have substituted her urine specimens with “clean” specimens between DAYS 6 and 13, inclusive. Fifth, she could have had high PLD concentrations or activity (enabling more efficiently PEth production from ethanol contained in the mouthwash) or a long PEth elimination half-life (27). If either or both of these were the case, however, it is not clear why the DAY 6 PEth concentration was not elevated from baseline. While the unexpected PEth increase in this participant was not inconsistent with our hypothesis, it highlights potential heterogeneity regarding individual differences in PEth response and suggests their consideration in future work.

This study has several limitations. Participants’ use of mouthwash was unwitnessed, and we did not employ objective methods to verify adherence with the study protocol (e.g., smartphone video of mouthwash use). However, we attempted to facilitate adherence in several ways. First, we chose mint-flavored Listerine^® in order to improve palatability relative to the original Listerine^® product. Second, we provided participants with prefilled, single-use mouthwash containers, thereby improving portability. Third, recognizing that participants might forget one or more gargling sessions during the day, we requested that they remedy any missed sessions with make-up sessions later in the day in order to accomplish four-times-daily use. We attempted to measure adherence at mid- and endpoints of the study by interviewing participants about their mouthwash use and by counting their empty single-use mouthwash containers.

We used self-reports and daily urinary EtG and EtS measurements to verify, ex post facto, participants’ reports of ethanol abstinence in the days prior to and during the study. Light ethanol consumption, however, could have gone undetected by measurement of urinary ethanol conjugates only. This issue was particularly relevant to Participant #6, who produced an unexplained PEth spike on the final study day, but denied ethanol consumption and produced no detectable urinary ethanol conjugates on any study day. Breath or transdermal ethanol monitoring, instead of, or as a complement to, urinary EtG/EtS monitoring, might have detected surreptitious ethanol consumption. However, both breath- and transdermal monitoring techniques have their own limitations in detecting low levels of ethanol consumption (28–30).

The study lasted 12 to 13 days. The published half-lives of PEth 16:0/18:1 range from ∼3.7 to 13.6 days (3, 11). Participants with half-lives at the longer end of this range, and who therefore failed to reach steady state, may have produced higher PEth concentrations during the course of a longer study.

Finally, the study did not include a control group. Although the current design was sufficient to address our hypothesis, inclusion of controls may have shed light on the relative contributions of mouthwash and irreducible extraneous ethanol exposures to the quantifiable PEth concentrations produced by some of the participants.

We have several suggestions for the design of future studies. First, as noted in the previous paragraph, inclusion of a control group could help explain the meaning of quantifiable but low PEth concentrations. Second, the use of breath or transdermal ethanol testing, to complement urinary ethanol conjugate testing, could improve the detection of surreptitious ethanol consumption. Third, a longer study period, perhaps up to 50 days, would allow the assessment of steady state PEth concentrations in individuals with long elimination half-lives. Fourth, measurement of PEth concentrations in advance of the start of the study might improve the possibility that participants would begin the study with nondetectable or very low PEth concentrations. Finally, future studies should address other important sources of extraneous ethanol exposure, including, in the midst of a viral pandemic, ethanol-containing hand sanitizer.

Conclusion

The results of this pilot study suggest that four-times-daily use of high-alcohol content mouthwash is unlikely to produce blood PEth concentrations that exceed the common commercially available threshold concentration of 20 ng/mL.

Funding

This study was supported by the Florida Recovery Center Pottash Research Initiative, Department of Psychiatry, University of Florida College of Medicine.

Conflict of interest

Joseph Jones is an employee of United States Drug Testing Laboratories, Inc., a commercial reference laboratory that is in the business of performing the tests discussed in this article.