Adolescents and Young Adults Have Difficulty Understanding Nicotine Concentration Labels on Vaping Products Presented as mg/mL and Percent Nicotine

Abstract

Introduction

E-cigarette e-liquid nicotine concentrations typically are labeled as mg/mL or percent nicotine. We examined whether these metrics accurately convey nicotine strength to young e-cigarette users and if youth can compare concentrations presented in mg/mL and percent nicotine.

Aims and Methods

Eight hundred and twenty-one adolescent and young adult e-cigarette users participated in the survey. Participants rated nicotine concentration strengths presented as mg/mL (0–60 mg/mL) and percent nicotine (0%–6%) from “no nicotine” to “very high nicotine.” Participants also viewed pairs of nicotine concentrations (eg, 18 mg/mL vs. 5%) and indicated which concentration was stronger or if the concentrations were equivalent.

Results

On average, participants correctly identified 5.92 (2.68) of 18 nicotine strengths, correctly identifying strengths labeled as mg/mL (3.47 [2.03]) more often than percent nicotine (2.45 [1.38], p < .001). Excluding nicotine-free, participants rated concentrations presented as mg/mL as stronger, more addictive, and more harmful than equivalent concentrations presented as percent nicotine. Participants seldom correctly identified that one concentration was stronger or that both were equivalent (7.58 [5.88] of 19 pairings), although they more often correctly identified the stronger concentration when it was presented in mg/mL (4.02 [SD = 3.01]) than in percent nicotine (2.53 [2.73], p < .001). The most consistent predictor of correct answers on these tasks was familiarity with using both products labeled as mg/mL and labeled as percent nicotine.

Conclusions

Young e-cigarette users had difficulty understanding nicotine concentrations labeled using the most common metrics, raising concerns about inadvertent exposure to high nicotine levels and suggesting that a more intuitive labeling approach is needed.

Implications

This study extends prior work showing that young e-cigarette users often are uncertain whether the e-liquids they use contain nicotine by demonstrating that adolescents and young adults have difficulty understanding nicotine concentrations labeled using the two most common metrics (mg/mL and percent nicotine). Errors generally underestimated nicotine strength, and users were not able to accurately compare nicotine concentrations presented as mg/mL and percent nicotine. Difficulty understanding labeling metrics persisted even after accounting for user characteristics like age and vaping experience, suggesting that a novel, easy to understand labeling system is needed to convey information about nicotine strength accurately.

Introduction

Since 2014, e-cigarettes have been the most commonly used nicotine product by adolescents,¹ and young adults ages 18–24 years have the highest prevalence of vaping among adults.^2,3 Recent data show that 27.5% of high school students currently vape,¹ and 5.2%³ to 9.2%² of young adults ages 18–24 vape. In 2016, the Surgeon General highlighted adolescent and young adult vaping as a public health crisis.⁴

Research suggests that vaping may be disproportionately harmful for adolescents and young adults (AYAs) relative to older e-cigarette users.⁵ Negative consequences include toxicant exposure,⁶ risk for combustible tobacco product use,^7–10 and exposure to nicotine,¹¹ which has been shown to have detrimental effects on the developing brain (eg, impairments in memory, impulse control, and executive functioning).⁴ Related, AYAs are highly vulnerable to developing nicotine dependence⁴ and can experience symptoms of nicotine dependence via vaping.¹² Despite the documented harms of nicotine exposure, many AYAs^13,14 are unsure whether the e-cigarette(s) they use even contain nicotine. This lack of basic knowledge about nicotine content suggests that knowledge of other important factors like nicotine concentration and addiction potential likely lags even further behind.

As of August 2018, the FDA requires e-cigarette packaging to display a standardized nicotine warning (“This product contains nicotine. Nicotine is an addictive chemical.”  ¹⁵). While the warning may improve AYAs’ knowledge of whether a product contains any nicotine,^16,17 it does not speak to the nicotine strength of a product. E-cigarette e-liquids and pods are available in a wide range of nicotine concentrations,¹⁸ which typically are labeled on packaging either as mg/mL or percent nicotine (eg, 50 mg/mL, 5%). The extent to which AYAs understand these metrics is uncertain, as comprehensive research has not been conducted on the topic. However, a recent study provides evidence that adolescents may misunderstand current labeling metrics. At the time of the study, JUUL e-cigarettes were labeled as containing 5% nicotine, which is a very high nicotine level. Importantly, when informed that JUULs contain 5% nicotine, most adolescents thought that 5% was a low nicotine level.¹⁹ As a result of this misunderstanding, youth inadvertently may be exposing themselves to high nicotine levels and increasing their risk for negative consequences like dependence.

In sum, given that AYAs constitute the largest percentage of vapers and are disproportionately likely to develop nicotine dependence, understandable labeling of nicotine strengths is needed to help young people appreciate the potential risks of different e-cigarette products and inform their decision making. The current study employed survey methodology to evaluate the extent to which high school- and college-aged AYAs understand nicotine concentrations presented using the two most common metrics: mg/mL and percent nicotine. Based on prior research indicating limited knowledge of nicotine content,^13,14 we anticipated that AYAs would have difficulty interpreting nicotine concentrations but that young adults may evidence a better understanding than adolescents given their older age and experience with vaping.

Methods

Participants

From October through December 2019, a nationwide sample of 821 lifetime e-cigarette users (44.7% high school students, 55.3% young adults) was recruited to complete a 20-minute online survey hosted via Qualtrics. Quotas were included to ensure sample diversity by sex, race, and region of the country. To ensure that young adults both attending and not attending college were recruited, a quota was set such that no more than 55% of the young adults were college students. Finally, a quota was set such that at least 50% of the total sample had to report vaping at least weekly to ensure that a sizeable portion had high familiarity with vaping.

Procedures

The Institutional Review Board of Oberlin College approved the study procedures. Participants were recruited directly through Qualtrics Online Sample, a secure, market research service operated by Qualtrics, Inc. Qualtrics Online Sample obtains parental consent for all underage participants who serve as panelists. Qualtrics sent email invitations to panelists who they deemed likely to be eligible for the study based on prior responses to a demographics survey(s) administered by Qualtrics at the time participants volunteered to become “panelists” (eg, lifetime vaping status). Interested individuals clicked on an embedded link, which directed them to the study eligibility questions (see Measures). To determine eligibility, participants reported on their age, enrollment in high school or college, sex, race, region of the country in which they reside, lifetime alcohol and cigarette use (which were used to disguise the purpose of the study), and lifetime e-cigarette use. Individuals who endorsed lifetime e-cigarette use reported whether they typically vape at least once a week (no, yes, I don’t know). Basic eligibility requirements included being 14–24 years old and reporting lifetime e-cigarette use. Participants meeting these criteria were eligible until quotas filled, at which point eligibility criteria shifted (eg, if female sex was full, only males were eligible). Eligible individuals provided assent (ages < 18 years) or consent (ages 18 years and older) prior to participating. Qualtrics directly compensated participants based on the terms of their pre-established agreements (not to exceed $8). Although the exact amount of compensation was not revealed to the authors, Qualtrics states that participants “receive an incentive based on the length of the survey, their specific panelist profile, and target acquisition difficulty. The specific types of rewards vary and may include cash, airline miles, gift cards, redeemable points, sweepstakes entrance, and vouchers” (European Society for Opinion and Market Research, 2014).

Primary Measures

Perceptions of Nicotine Strength

AYAs answered questions about the perceived strengths of concentrations presented as mg/mL and as percent nicotine, respectively, in two randomly presented blocks. The order of presentation of nicotine concentrations was randomized within each block. In total, AYAs rated 18 nicotine concentrations (0 mg/mL, 0%; 3 mg/mL, 0.3%; 6 mg/mL, 0.6%; 18 mg/mL, 1.8%; 30 mg/mL, 3%; 40 mg/mL, 4%; 45 mg/mL, 4%; 50 mg/mL, 5%; 60 mg/mL, 6%). For each concentration, participants selected from seven nicotine strength response options (eg, 6 mg/mL of nicotine is… [nicotine free, very low, low, medium, high, very high, I don’t know]). The nicotine strengths asked about and the response options provided were based on formative work and ratings by 40 experts in the e-cigarette field during instrument development (Supplement A).

For each nicotine concentration, AYAs also reported “how addictive [they] think it is to vape e-liquid containing [concentration] of nicotine?” and “how harmful [they] think it is to vape e-liquid containing [concentration] of nicotine?” (not at all, slightly, moderately, very, extremely). The order of presentation of these two questions was randomized.

Head-to-Head Comparisons of Nicotine Strengths Presented as mg/mL and Percent Nicotine

AYAs viewed 25 pairs of nicotine concentrations (eg, 18 mg/mL vs. 5%) and were asked to indicate which concentration was stronger or if the two concentrations were equivalent. AYAs also could indicate that they did not know if or how the two concentrations differ. There were eight pairings in which the concentration presented as mg/mL was stronger than the percent nicotine concentration (eg, 36 mg/mL vs. 3%), eight pairings in which percent nicotine was stronger than mg/mL (eg, 24 mg/mL vs. 3%), and three cases in which the concentrations presented as mg/mL and percent nicotine were equal (eg, 50 mg/mL vs. 5%). Six pairings also presented concentrations using the same nicotine metric (eg, 6 vs. 12 mg/mL, 3% vs. 5%) to ensure that AYAs understood the task and were paying attention. The 25 pairs were presented in a randomized order.

Secondary Measures

Awareness of Nicotine Concentration Labels

AYAs responded to the following: “Currently, the nicotine concentration (or strength) of e-liquids and/or pods for vaping devices is printed on the product packaging” (response options: false, true, I don’t know). Data were coded as awareness (no/yes).

E-cigarette Use

AYAs reported on age at vaping onset and lifetime use of five e-cigarette device types: disposable vapes or cig-a-likes, vape pens, JUUL, pod devices other than JUUL, and mods or APVs. For each device an AYA reported ever using, they reported on frequency of use in the past 30 days (0–30 days, from which weekly use of each device was calculated) and whether they typically vape nicotine in the product (no, yes, I don’t know). If an AYA indicated nicotine use, they reported how the nicotine concentration typically is labeled on the product they use (mg/mL, percent nicotine, I don’t know). While some endorsed using only one device with nicotine labeled either as mg/mL or percent nicotine, many AYAs indicated using multiple devices, often with different nicotine concentration labels (eg, JUUL use with nicotine labeled at percent nicotine and a vape-pen with nicotine labeled as mg/mL), and a considerable number did not know how the nicotine concentration on the product(s) they use was labeled. As such, a variable was created reflecting global familiarity with nicotine concentration labels presented as mg/mL and percent nicotine (coded as neither, mg/mL only, percent nicotine only, and both mg/mL and percent nicotine). Finally, AYAs reported on device ownership (Whose vaping device do you use [or did you use]? Select all that apply from: my own device, my friends’ [or share with friends], my parents’, other [write-in]). Responses were coded as device ownership (no or yes).

Analytic Plan

Descriptive Statistics

Descriptive statistics were run on all study variables. Chi-square analyses (for categorical variables) and ANOVAs (for continuous variables) were run to evaluate potential differences on study variables by age group.

Perceived Nicotine Strength

First the mean numbers of “correct” classifications were calculated for all 18 nicotine concentrations and by mg/mL (n = 9) and percent nicotine (n = 9). We defined “correct” classifications as those in which perceived nicotine strength for a given concentration aligned with the strength category endorsed by the majority of experts in the field (eg, an answer of “high” for 30 mg/mL was deemed “correct” based on consistency with expert opinion; average consistency across concentration strength ratings for experts = 74.9%). All other responses were marked as “incorrect.” Given that there was variability in subject matter experts’ opinions, especially for intermediate concentrations (eg, 18 mg/mL; Supplement A), we also calculated correct and incorrect scores based off of the two most commonly endorsed strength categories. For example, 55.6% of experts rated 30 mg/mL nicotine as “high” and 24.2% rated 30 mg/mL as “very high” so we considered ratings of high or very high. This resulted in an average consistency across expert ratings of 91.7%. While this approach resulted in more correct responses among youth (eg, total correct responses using one category = 5.92 [2.68]; using two categories = 8.49 [3.44]), the findings mirrored those reported below. To simplify the presentation of results, we only present findings in which “correct” responses were linked to the most commonly endorsed strength category for each concentration.

Given that there were no differences in the number of correct responses for strength ratings by age status (ie, high school vs. college) (see Results), a paired-samples t test was run within the total sample to evaluate whether AYA vapers had more difficulty understanding concentrations presented as mg/mL or percent nicotine. The overall percentages of correct responses, responses that overestimated the strength (relative to expert rating), and responses that underestimated nicotine strength, were also examined. Lastly, given that device type can impact nicotine delivery,²⁰ the impact of using different devices (eg, JUULs, mods) on perceptions of nicotine strength was examined.

Finally, three univariate general linear models were run to examine predictors of the number of correct strength ratings for each concentration type (mg/mL, percent nicotine) and overall. Independent variables included demographic characteristics (age status, sex, race); age at vaping onset, device ownership (no, yes); knowledge that e-liquid and pod packaging includes a nicotine concentration label (no/yes); familiarity with using e-liquids that are labeled using different nicotine concentration labels (coded as familiarity with neither mg/mL nor percent nicotine, mg/mL only, percent nicotine only, and both mg/mL and percent nicotine), and weekly use of each device type (no or yes). Models initially were run with two-way interactions between high school and college status. Given that two-way interactions were not significant (results not presented), main effects models were run to increase parsimony.

Perceived Nicotine Strength, Addictive Potential, and Overall Harm of Nicotine Concentrations

Paired-samples t tests were used to examine mean differences in AYAs’ perceptions of the strength, addiction potential, and overall harm associated with the range of corresponding nicotine concentrations presented as mg/mL and percent nicotine (eg, 3 mg/mL vs. 0.3%).

Head-to-Head Comparisons of Nicotine Strengths

The mean numbers of “correct” answers for different pairing presentation types were calculated and ANOVAs were run to evaluate whether the total numbers of correct answers differed by high school or college status. Given that no differences were observed by age status, a paired-samples t test was run within the total sample to evaluate whether young vapers were more likely to identify “correctly” the stronger nicotine concentration when the correct answer was presented as mg/mL or percent nicotine.

Finally, four univariate GLM models were run to examine predictors of the number of correctly identified head-to-head comparisons (total, mg/mL stronger than percent nicotine, percent nicotine stronger than mg/mL, mg/mL and percent nicotine equivalent). The independent variables mirrored those described above for nicotine strength. Again, two-way interactions between independent variables and high school and college status were not significant, so main effects models are presented.

Results

Sample Description

Descriptive statistics for are reported in Table 1. Compared with high school students, college-age individuals (not enrolled or enrolled in college) were more likely to start vaping at a later age; have a racial background other than non-Hispanic White or Black; own their own vaping device; engage in weekly use of cig-a-likes or disposables; vape pens, and/or mods; and be aware that nicotine concentrations are labeled on e-liquid and pod packaging. Young adults who were not enrolled in college were most likely to be familiar with nicotine concentrations presented only as mg/mL, while those in college were most likely to use a greater total number of vaping devices weekly and to be familiar with nicotine concentrations presented both in mg/mL and percent nicotine.

Table 1.

Descriptive Statistics and Differences on Study Variables by Age Status

	Total sample	High school	College (not enrolled)	College (enrolled)
	N = 821	n = 367	n = 222	n = 232
Sex (% female)	60.0	60.2 A	61.3 A	61.3 A
Age	18.95 (2.81)	16.51 (1.25) A	21.21 (2.10) B	20.63 (2.03) B
Age vaping onset	16.61 (2.41)	15.02 (1.33) A	17.98 (2.44) B	17.79 (2.21) C
Race
White	64.9	71.7 A	62.2 B	56.9 B
Black	15.8	16.9 A	13.1 A	16.8 A
Other	19.2	11.4 A	24.8 B	26.3 B
Region of country
Northeast	20.3	21.8 A	15.3 A	22.8 A
Midwest	21.8	21.3 A	27.0 A	17.7 A
South	38.2	34.9 A	41.0 A	40.9 A
West	19.6	22.1 A	16.7 A	18.5 A
Own device	54.0	44.1 A	60.8 B	62.9 B
Weekly use of any device (yes)	56.9	50.7 A	60.4 B	63.4 B
Total devices (weekly use)	1.43 (1.57)	1.24 (1.50) A	1.54 (1.56) A	1.64 (1.65) B
Cigalike or disposable	28.0	23.2 A	32.0 B	31.9 B
Vape-pen	31.8	26.2 A	35.1 B	37.5 B
JUUL	35.7	32.4 A	36.0 A	40.5 A
Other pod devices	27.9	30.0 A	23.9 A	28.4 A
Mods	20.1	12.3 A	27.0 B	25.9 B
Awareness nicotine concentration	66.7	58.3 A	73.0 B	74.1 B
Familiarity with nicotine concentrations
None	26.9	34.6 A	18.9 B	22.4 B
mg/mL only	19.6	12.8 A	32.0 B	18.5 A
Percent nicotine only	28.0	29.2 A	26.1 A	28.0 A
Both mg/mL and percent nicotine	25.5	23.4 A	23.0 A	31.0 B
Total correct nicotine strength ratings (of 18)	5.92 (2.68)	5.74 (2.76) A	6.05 (2.60) A	6.09 (2.65) A
mg/mL (of 9)	3.47 (2.03)	3.29 (1.97) A	3.59 (2.06) A	3.63 (2.08) A
Percent (of 9)	2.45 (1.38)	2.45 (1.46) A	2.45 (1.25) A	2.45 (1.36) A
Total correct nicotine comparisons (of 19)	7.58 (5.88)	7.87 (6.13) A	7.18 (5.88) A	7.52 (5.47) A
mg/mL stronger (of 8)	4.02 (3.02)	4.07 (3.03) A	3.93 (3.03) A	4.05 (2.99) A
Percent stronger (of 8)	2.53 (2.73)	2.68 (2.79) A	2.27 (2.72) A	2.54 (2.65) A
mg/mL and percent are equal (of 3)	1.03 (1.27)	1.13 (1.30) A	0.98 (1.26) A	0.93 (1.23) A
mg/mL vs. mg/mL or percent vs. percent (of 6)	5.05 (1.78)	5.06 (1.82) A	5.14 (1.70) A	4.99 (1.80) A

Within rows, differing letters (denoted by A, B, and C) indicate a significant difference based on ANOVA or chi-square. To aid in interpretation, if all values in a row are accompanied by an A, there is no significant difference between groups. If values in a row are labeled as A, B, B, then groups two and three differ from group 1 but do not differ from each other. If values are labeled as A, B, C within a row, each group differs from the other groups.

Perceived Nicotine Strength

On average, AYAs correctly identified 5.92 (2.68) nicotine strengths out of 18. AYAs correctly identified the strengths of concentrations presented as mg/mL (3.47 [SD = 2.03]) more often than for percent nicotine (2.45 [1.37]), t(820) = 13.23, p < .001).

With regard to the percentage of correct and incorrect responses (Table 2), when concentration was presented as percent nicotine compared with mg/mL, AYAs were less likely to indicate a correct strength rating (percent 27.3%; mg/mL 38.5%), to say “I don’t know” (percent 10.3%; mg/mL 17.2%), or to overestimate nicotine strength (percent 3.0%; mg/mL 7.2%) and were more likely to underestimate nicotine strength (percent 59.4%; mg/mL 36.9%). The same pattern of results was observed, although errors were more pronounced, when “I don’t know” responses were excluded (correct answers [percent 30.4%; mg/mL 46.6%]; overestimating nicotine strength [percent 3.4%; mg/mL 8.9%]; underestimating nicotine strength [percent 66.2%; mg/mL 44.5%]).

Table 2.

Predictors of Correct Nicotine Strength Ratings

	Total correct nicotine concentration strength ratings
	All concentrations			mg/mL			Percent nicotine
	Adjusted R2 = 0.06			Adjusted R2 = 0.10			Adjusted R2 = 0.03
	B	Std. error	$n_p^2$	B	Std. error	$n_p^2$	B	Std. error	$n_p^2$
Intercept	6.58***	0.77	0.08	3.91***	0.57	0.06	2.67***	0.40	0.05
Age status
College students	0.39	0.26	0.00	0.35	0.19	0.00	0.04	0.14	0.00
College age (not enrolled)	0.30	0.27	0.00	0.20	0.20	0.00	0.11	0.14	0.00
High school age (ref)
Male sex	0.08	0.19	0.00	−0.07	0.14	0.00	0.15	0.98	0.00
Race
Other race	0.05	0.24	0.00	−0.09	0.18	0.00	0.14	0.13	0.00
Black	−0.21	0.26	0.00	0.02	0.20	0.00	−0.23	0.14	0.00
Non-Hispanic White (ref)
Age at vaping onset	−0.12	0.05	0.01	−0.09	0.04	0.01	−0.03	0.25	0.00
Device ownership (yes)	0.30	0.21	0.00	0.25	0.15	0.00	0.05	0.11	0.00
Aware of concentration on packaging	0.21	0.22	0.00	0.24	0.16	0.00	−0.04	0.11	0.00
Familiarity with nicotine concentrations+
Both mg/mL and percent	1.48***	0.30	0.03	1.04***	0.22	0.03	0.44**	0.16	0.01
mg/mL only	1.37***	0.30	0.03	1.50***	0.22	0.06	−0.14	0.16	0.00
Percent only	0.67	0.27	0.01	0.29	0.20	0.00	0.38**	0.14	0.01
None (ref)
Weekly device use
Cig-a-like or disposable use	−0.08	0.25	0.00	−0.02	0.18	0.00	−0.05	0.13	0.00
Vape-pen use	−0.27	0.24	0.00	−0.24	0.18	0.00	−0.04	0.13	0.00
JUUL use	0.00	0.24	0.00	−0.14	0.18	0.00	0.14	0.13	0.00
Pod use	−0.07	0.24	0.00	−0.05	0.18	0.00	−0.02	0.13	0.00
Mod use	0.27	0.26	0.00	0.31	0.19	0.00	−0.03	0.13	0.00

A Bonferroni adjusted alpha value of 0.017 was used to adjust for the fact that three models were run.

**p < .01, ***p < .001.

⁺Simple effects showed that having familiarity with concentrations presented both as mg/mL and percent nicotine was associated with more correct categorizations overall (compared with familiarity with neither mg/mL nor percent nicotine and familiarity only with percent nicotine), for mg/mL (compared with familiarity with neither mg/mL nor percent nicotine and familiarity only with percent nicotine), and for percent nicotine (compared with familiarity with neither mg/mL nor percent nicotine and familiarity only with mg/mL). Further, having familiarity only with concentrations presented in mg/mL was associated with more correct categorizations for mg/mL (compared with familiarity with neither mg/mL nor percent nicotine). Finally, familiarity only with percent nicotine was associated with more correct categorizations for percent nicotine (compared with familiarity with neither mg/mL nor percent nicotine and mg/mL only).

In adjusted regression models (Table 2), only preexisting familiarity with nicotine concentrations was significantly associated with higher numbers of correct nicotine strength ratings.

Perceived Nicotine Strength, Addictive Potential, and Overall Harm of Nicotine Concentrations

With the exception of nicotine free concentrations (0 mg/mL and 0%), AYAs were more likely to think that all concentrations presented as mg/mL were stronger, more addictive, and more harmful than corresponding concentrations presented as percent nicotine (p values < .001; Table 3).

Table 3.

Differences in Perceived Nicotine Strength, Addictive Potential, and Overall Harm When Concentrations Are Presented as mg/mL and Percent Nicotine

Concentrations	Strength				Addictive potential				Overall harm
	mg/mL		Percent nicotine		mg/mL		Percent nicotine		mg/mL		Percent nicotine
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
0 mg/mL; 0%	1.12	0.61	1.08	0.49	1.63	1.30	1.58	1.26	1.92	1.37	1.89	1.36
3 mg/mL; 0.3%	2.44	0.79	2.17	0.67	3.05	1.52	2.67	1.43	3.01	1.52	2.64	1.39
6 mg/mL; 0.6%	2.90	1.04	2.33	0.78	3.35	1.48	2.78	1.42	3.25	1.48	2.77	1.40
18 mg/mL; 1.8%	3.59	0.96	2.68	0.87	3.69	1.37	3.08	1.37	3.60	1.39	3.00	1.31
30 mg/mL; 3.0%	4.25	0.89	3.16	1.02	3.96	1.24	3.27	1.29	3.80	1.27	3.18	1.30
40 mg/mL; 4.0%	4.74	0.85	3.56	1.16	4.21	1.13	3.51	1.31	4.01	1.18	3.37	1.28
45 mg/mL; 4.5%	4.86	0.84	3.80	1.17	4.25	1.12	3.64	1.28	4.05	1.20	3.49	1.27
50 mg/mL; 5.0%	5.15	0.87	3.94	1.29	4.38	1.07	3.71	1.29	4.21	1.15	3.52	1.25
60 mg/mL; 6.0%	5.50	0.73	4.20	1.39	4.56	0.98	3.86	1.27	4.37	1.10	3.67	1.27

Bolded values indicate significant differences at p < .001. Response options for strength: 1 (nicotine free), 2 (very low), 3 (low), 4 (medium), 5 (high), 6 (very high); response options for addictive potential and overall harm: 1 (not at all), 2 (slightly), 3 (moderately), 4 (very), 5 (extremely).

Head-to-Head Comparisons of Nicotine Strengths

When comparing strengths that were presented using the same metric (eg, 12 vs. 30 mg/mL), AYAs generally performed well (M = 5.05 [SD =1.78] correct answers out of six comparisons), indicating that they understood the task and were paying attention. However, when concentrations in a pairing were presented as mg/mL versus percent nicotine (eg, 12 mg/mL vs. 2.4%), AYAs were less accurate at identifying which concentration was stronger or that both were equivalent (M = 7.58 [5.88] correct out of 19 pairings). Although the total numbers of correct responses were low, AYAs correctly identified the stronger nicotine concentration when the correct answer was presented in mg/mL (4.02 [SD = 3.01]) more often than when the correct answer was in percent nicotine (2.53 [2.73]), t(820) = 14.81, p < .001).

For the regression models predicting correct responses (see Table 4), a Bonferroni adjusted alpha value of 0.0125 was used to account for the fact that four models were run. Again, preexisting familiarity with nicotine concentrations was significantly associated with more correct answers. Being male and being aware that nicotine concentrations are labeled on e-cigarette e-liquid and pod packaging were associated with more correct responses for each of the four models. Finally, using disposable e-cigarettes weekly was associated with more correct answers when percent nicotine was stronger than mg/mL, likely due to the fact that many of these are labeled using percent nicotine.

Table 4.

Total Correct Head-to-Head Comparisons of Nicotine Concentrations Presented as mg/mL and Percent Nicotine

	Total correct head-to-head comparisons of nicotine concentrations presented as mg/mL and percent nicotine
	All comparisons			When mg/mL is stronger			When percent nicotine is stronger			When mg/mL and percent are equal
	Adjusted R2 = 0.14			Adjusted R2 = 0.11			Adjusted R2 = 0.10			Adjusted R2 = 0.09
	B	Std. error	$n_p^2$	B	Std. error	$n_p^2$	B	Std. error	$n_p^2$	B	Std. error	$n_p^2$
Intercept	7.12***	1.61	0.02	3.09***	0.84	0.02	3.01***	0.76	0.02	1.02**	0.36	0.01
Age status
College students	−0.53	0.55	0.00	−0.31	0.29	0.00	−0.04	0.26	0.00	−0.18	0.12	0.00
College age (not enrolled)	−0.70	0.57	0.00	−0.40	0.30	0.00	−0.22	0.27	0.00	−0.07	0.13	0.00
High school age (ref)
Male sex	1.87***	0.40	0.03	0.67***	0.21	0.01	0.96***	0.19	0.03	0.24**	0.09	0.01
Race
Other race	1.04	0.51	0.01	0.69	0.27	0.01	0.25	0.24	0.00	0.10	0.11	0.00
Black	0.99	0.55	0.00	0.68	0.29	0.01	0.18	0.26	0.00	0.14	0.12	0.00
Non-Hispanic White (ref)
Age at vaping onset	−0.19	0.10	0.00	−0.05	0.05	0.00	−0.11	0.05	0.01	−0.03	0.02	0.00
Device ownership (yes)	−0.15	0.43	0.00	0.19	0.22	0.00	−0.18	0.20	0.00	−0.16	0.10	0.00
Aware of concentration on packaging	1.83***	0.46	0.02	0.94***	0.24	0.02	0.59**	0.22	0.01	0.30**	0.10	0.01
Familiarity with nicotine concentrations+
Both mg/mL and %	3.14***	0.63	0.03	1.55***	0.33	0.03	0.97***	0.30	0.01	0.62***	0.14	0.02
mg/mL only	1.64**	0.63	0.01	1.30***	0.33	0.02	0.34	0.30	0.00	0.10	0.14	0.00
Percent only	0.82	0.57	0.00	0.29	0.30	0.00	0.39	0.27	0.00	0.14	0.13	0.00
None (ref)
Weekly device use
Cig-a-like or disposable use	1.16	0.52	0.01	0.09	0.27	0.00	0.84***	0.25	0.01	0.22	0.12	0.01
Vape-pen use	−1.03	0.51	0.01	−0.27	0.26	0.00	−0.55	0.24	0.01	−0.21	0.11	0.00
JUUL use	0.21	0.50	0.00	−0.04	0.26	0.00	0.19	0.24	0.00	0.07	0.11	0.00
Pod use	0.88	0.51	0.00	0.38	0.27	0.00	0.24	0.24	0.00	0.26	0.11	0.01
Mod use	−0.14	0.54	0.00	0.01	0.28	0.00	−0.06	0.26	0.00	−0.08	0.12	0.00

*p < .05, **p < .01, ***p < .001.

⁺Simple effects showed that having familiarity with concentrations presented both as mg/mL and percent nicotine (compared with neither or percent only) was associated with more correct answers overall, when mg/mL was stronger than percent nicotine (compared with neither or percent only), when percent nicotine was stronger than mg/mL (compared with familiarity with neither), and when mg/mL and percent nicotine were equal (compared with familiarity with mg/mL only, percent only, or neither). Familiarity with nicotine concentrations presented only as mg/mL was associated with more correct answers overall (compared with percent nicotine only or neither) and with more correct answers when mg/mL was stronger than percent nicotine (compared with percent only and neither).

Discussion

The current study is the first, to our knowledge, to assess directly the extent to which young e-cigarette users understand nicotine concentrations labeled using the two most common metrics: mg/mL and percent nicotine. In summary, study results indicated that current labeling metrics do not convey information adequately about nicotine strength to AYAs and that AYAs are not meaningfully able to compare nicotine concentrations presented using the two different metrics.

Regarding nicotine strength, AYAs frequently were inaccurate when identifying the strength of nicotine concentrations presented as mg/mL and percent nicotine. While they performed slightly better when nicotine concentrations were presented as mg/mL versus percent nicotine, on average AYAs’ ratings of the strength of nicotine concentrations were only consistent with experts’ rating about one third of the time. This result suggests that nicotine concentrations on e-cigarette e-liquid and pod labels are not communicating information effectively regarding nicotine strength to young users, who constitute the largest percentage of vapers and a highly vulnerable population.^1–4 Further, when AYAs incorrectly identified nicotine strengths, on average, they were more likely to underestimate than to overestimate strength. This underestimation raises concerns about inadvertent exposure to high levels of nicotine via misunderstanding of current nicotine concentration labels. The fact that the prevalence of underestimating nicotine strength was magnified further when concentrations were presented as percent nicotine, coupled with evidence that youth also perceived concentrations presented as percent nicotine as less addictive and less harmful, is particularly concerning given the popularity of pod devices like JUUL among youth²¹ which typically carry labels in percent nicotine.²² Finally, the models predicting the mean number of “correct” nicotine strength classifications revealed little additional information. Only familiarity with nicotine concentrations (both mg/mL, mg/mL only, and percent nicotine only) was significantly associated with correct strength ratings across the models, with modest effect sizes (1%–6%). Neither being college-aged (vs. high school-aged), sex, race, age at vaping onset, device ownership, awareness of nicotine warnings, nor weekly device use was significantly associated with the total numbers of correct strength ratings. Although the lack of findings could be interpreted a number of ways, one possible explanation is that misunderstanding of nicotine strength is widespread among AYAs irrespective of many characteristics that reflect seemingly important constructs related to vaping experience (eg, device ownership, duration of use, use of specific devices, age or education level).

In addition to pervasive misunderstanding of nicotine strength, AYAs were inaccurate when comparing nicotine concentrations presented as mg/mL and percent nicotine, suggesting that a single, easy to understand method for labeling nicotine concentrations may be warranted. AYAs were able to identify correctly that one concentration was stronger than the other (eg, 18 mg/mL vs. 2.4%) or that the concentrations were equivalent (eg, 18 mg/mL vs. 1.8%), on average, less than 50% of the time (M = 7.58 [5.88] of all 19 pairings). Again, AYAs were more likely to correctly identify the stronger nicotine concentration when the correct answer was presented in mg/mL. However, this may have reflected a response based on numerical value rather than on a true understanding of nicotine concentration; in each case the numerical value associated with each concentration presented as mg/mL was higher than numerical value of the percentage (eg, 18 mg/mL vs. 2.4%). Further, while AYAs performed worse on both tasks when nicotine was presented as percent nicotine, this result may have little practical meaning given the overall low rates of correct responses. Finally, results from the adjusted models provided modest incremental information. Significant predictors of “correct” responses largely were driven by familiarity with nicotine concentrations (in this case, familiarity with labels as mg/mL and percent nicotine or general awareness that e-liquid and pod packaging carries nicotine concentration labels). The only other consistent finding was that males were more likely than females to provide more correct answers on the task assessing head-to-head comparisons of nicotine concentration, although it is not clear exactly why this may be.

The study findings should be considered in light of several limitations. Although quotas were put in place to ensure diversity of the sample by age status, college enrollment, race, region of the county, and e-cigarette use frequency, the generalizability of the findings may be limited by the use of a sample comprising Qualtrics panelists. Further, given that many AYAs used multiple devices with equal frequency, many of which had packaging with different types of nicotine concentration labels, isolating the impact of single device use and familiarity with a single method of labeling nicotine concentrations was not possible. Although this limitation adds some statistical noise, it also may reflect accurately the way in which youth use these devices. These data also did not differentiate freebase nicotine from nicotine salts, the latter of which is associated with increased nicotine delivery and risk for addiction.²³ In addition, no context explicitly was provided to participants regarding how to interpret strength (eg, strength relative to all commercially available nicotine concentrations, relative to all possible concentrations [0%–100%], or relative to an anchoring product like a combustible cigarette). As such, it is not possible to know what metric participants used when categorizing nicotine concentration strengths. Related, our operationalization of “correct answers” for perceptions of nicotine strength (eg, low, medium, high) was based on the opinions of experts who conduct research in the e-cigarette field. Given that there was disagreement regarding how best to categorize nicotine concentration strengths into simplified categories, future research is needed to determine the best metric against which to anchor nicotine concentration strengths. Specifically, it will be critical to determine how best to map labels (eg, low, medium, high) onto a more objective index or indices of strength and/or addictive potential (eg, the amount of nicotine that is delivered relative to a combustible cigarette; the relative increase in addictive potential associated with using higher levels of nicotine). Ideally, this anchor will optimize the utility of the labels for dissuading AYAs who have never tried vaping from doing so, encouraging AYAs who are experimenting to avoid using nicotine concentrations that are more likely to quickly produce nicotine addiction, and helping AYAs who currently smoke cigarettes and wish to switch to vaping select an appropriate nicotine concentration to facilitate a successful switch.

In conclusion, the current study provides evidence that high school- and college-aged AYAs do not understand how to interpret the strength of nicotine concentrations of vaping products labeled as mg/mL or percent nicotine. Those products labeled as percent nicotine may be even more problematic in leading to underestimation of nicotine strength. Further, AYAs appear to be unable to meaningfully compare nicotine concentrations presented using these two metrics. In sum, the findings suggest that efforts to develop a singular, simplified, and interpretable method for labeling the nicotine concentrations of vaping products that is linked to an objective anchor is needed to better inform and potentially protect vulnerable young e-cigarette users. Given that the US Food and Drug Administration has authority to regulate marketing and labeling practices, development and adoption of such a novel, interpretable labeling system should be prioritized.

Supplementary Material

A Contributorship Form detailing each author’s specific involvement with this content, as well as any supplementary data, are available online at https://academic.oup.com/ntr.

Funding

This project was supported by the Tobacco Centers of Regulatory Science (TCORS) award U54CA180908 from the National Cancer Institute (NCI) and Food and Drug Administration (FDA) Center for Tobacco Products (CTP). Dr Eissenberg’s effort is supported by grant number U54DA036105 from the National Institute on Drug Abuse of the National Institutes of Health and CTP/FDA. Efforts by Drs Wackowski and Delnevo are supported in part by grant R37CA222002 from NCI and U54CA229973 from the FDA and NCI. Dr Krishnan-Sarin’s effort is supported by grant number U54DA036151 from the National Institute on Drug Abuse of the National Institutes of Health and CTP/FDA. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the FDA.

Declaration of Interests

Dr Eissenberg is a paid consultant in litigation against the tobacco industry and also the electronic cigarette industry and is named on a patent for a device that measures the puffing behavior of electronic cigarette users. Dr Morean holds a restricted stock agreement with Gofire, Inc. although this is not directly relevant to the current study. Drs Wackoski, Delnevo, and Krishnan-Sarin have no conflicts of interest to declare related to the current study.