Impact of the California State Flavored Tobacco Sales Restriction on E-cigarette Use Behaviors Among Youth

Abstract

Background:

In December 2022, California prohibited the sale of most flavored tobacco products; however, limited data are available assessing the impact on adolescent tobacco use behaviors.

Methods:

We compared data from youth in 11^th grade prior to (fall 2022; n=1212; “pre-law”) and after implementation of the law (fall 2023; n=1026; “post-law”) at 4 schools in Southern California (panel study). We assessed past 30-day e-cigarette use, e-cigarette flavors used, perceived difficulty in getting preferred flavors, and (among new initiators), flavor at first use, and support for, anticipated impact, and perceived actual impact of the law. We also assessed change in e-cigarette use behaviors pre-law, pre- to post-law, and post-law in the same individuals (cohort study).

Results:

There was little change in past 30-day e-cigarette use pre-law (5.8%) vs. post-law (4.1%) (p=0.11). At both timepoints, all new initiators (100%) reported starting vaping with a flavored product, most reported past 30-day flavored e-cigarette use (89.8% vs. 90.9%), and most also reported that access to flavors was very or somewhat easy (85.5% vs. 83.9%). Youth observed limited impact on e-cigarette use among peers. In the cohort study, few (6.3%) switched completely from e-cigarettes to another nicotine product; many (46.5%) reported no past 30-day nicotine use post-law (demonstrating a similar pattern as pre-law).

Conclusions:

We observed limited change in the use of flavored e-cigarettes after a state-wide law restricting flavored tobacco sales was enacted. Findings may reflect inadequate implementation (e.g., insufficient enforcement efforts), which warrants further study to reduce youth access to flavored tobacco products.

INTRODUCTION

Although e-cigarette use has continued to decline among youth in the United States (US) over the last 4 years, e-cigarettes remain one of the most used substances among young people. In 2024, the national prevalence of past-year e-cigarette use among 12^th-grade students was 21.0%, and the prevalence of past 30-day use was 15.0%.^1,2 Most adolescents who use e-cigarettes regularly report that the first e-cigarette they used was flavored, and currently report using flavored e-cigarettes exclusively.^3,4 Flavors are a characteristic central to tobacco product appeal overall,⁵ particularly for e-cigarettes, and the industry has fought hard to limit U.S. Food and Drug Administration (FDA) regulatory authority over flavors.⁶ Flavors are a key reason for nicotine use among youth,⁷ and a substantial proportion of adolescents have reported that they would no longer use e-cigarettes if they were not available in flavors.^8,9

Flavors increase the abuse liability of nicotine products. Determinants of abuse liability – defined as the risk of persistent or problematic use of a drug, and the risk of undesirable consequences of use – include nicotine exposure, positive sensory effects, marketing, perceptions of harm, and social acceptance.¹⁰ Flavors, especially menthol or other cooling agents, reduce the unpleasant initial sensory effects of nicotine and increase the positive sensory effects associated with use.^11,12 This is an important determinant of progression to regular use and dependence.^13,14 We previously reported that first use of e-cigarettes in sweet flavors or mint/menthol (vs. tobacco flavor) was associated with a greater positive reaction to e-cigarette use; this was, in turn, associated with greater odds of continued vaping, greater number of vaping days in the past 30 days, and greater odds of nicotine dependence.¹⁵

Multiple policies have been proposed at the federal and state levels to reduce the use of flavored tobacco products among young people. The goals of these policies are two-fold: 1) prevent initiation among young people (especially those with no prior history of nicotine use); and 2) encourage cessation among those who are currently using flavored nicotine products. In California, Senate Bill (SB) 793 came into effect in December 2022, despite strong tobacco industry opposition, prohibiting the sale of most flavored tobacco products in the state (notably, after amendment, exemptions were made for flavored hookah, roll-your-own tobacco, and premium cigars).^16,17

Existing studies that have attempted to evaluate the efficacy of prohibitions on flavored tobacco products have largely been subject to limitations in the timing of data collection (e.g., using retrospective recall of behaviors), or have relied on other non-behavioral methods of assessment (e.g., product sales, or studies of packaging litter) to estimate the availability of flavored products and use behaviors of young people following implementation of the law. One study of California adolescents aged 12-17 evaluated the prevalence of past 30-day flavored tobacco product use prior to the law compared to after the law was enacted among two independent groups of youth assessed cross-sectionally, and reported widespread access to and use of flavored tobacco products even after the law was effective.¹⁸ The literature could be strengthened by: (a) assessing these research questions in cohort studies with multiple waves of data collection from the same participants in order to assess flavors used at recent initiation (e.g., in the prior 6 months), which reduces recall bias and ensures initiation took place during a short defined period of time with consistent policies in place), and (b) using prospective data from individuals who used flavored e-cigarettes prior to implementation of the law to assess the change in subsequent use behaviors.

In the current study, we examined patterns of adolescent e-cigarette use among 11^th graders in the same schools in Southern California before the state law restricting sales of most flavored tobacco products was implemented (fall 2022) and after the policy was in place (fall 2023). We sought to assess changes in (a) e-cigarette initiation with flavored products and (b) past 30-day use of flavored e-cigarettes prior to vs. after implementation of the state law. We also aimed to assess changes in source (for obtaining e-cigarettes), perceptions of peer use, and perceived ease of access to flavored tobacco products. Finally, we examined patterns of nicotine use behaviors post-law among those who reported past 30-day e-cigarette use pre-law to determine the extent to which youth were switching to other products or flavors because of the state law.

METHODS

Study Design and Participant Sample

Data were from an ongoing, split prospective cohort study of participants initially recruited in 9^th grade in either fall 2020 (Class of 2024) or fall 2021 (Class of 2025), as described previously.¹⁹ The cohort follows students from 9^th grade through 12^th grade, collecting data every semester. Because of the split cohort design, at any given wave of data collection, participants include students from two grade levels (e.g., in fall 2022, participants were in 10^th or 11^th grade; in fall 2023, participants were in 11^th or 12^th grade). Briefly, schools were recruited to maximize differences by sociodemographic characteristics, with a focus on recruiting schools representing individuals from various racial/ethnic backgrounds, a range of socioeconomic levels, and variability in terms of urban and suburban locations. Participants completed surveys every 6 months (once per semester) administered online via REDCap; beginning in fall 2021, all students completed surveys during class time at their schools with study staff onsite. Students who were not in class were sent a link and invited to complete the survey outside of class time.

We leveraged both the split cohort component and the prospective component of this study design in order to investigate two key research questions (see Supplemental Table 1). First, we utilized the split cohort component to examine how e-cigarette use behaviors differed for students who were in 11^th grade prior to the law (fall 2022), compared to those who were in 11^th grade after the law was enacted (fall 2023) (hereafter referred to as the “panel study”). Then, we utilized the prospective cohort component to examine how e-cigarette use behaviors changed from the wave prior to the law (fall 2022) to the wave after the law (fall 2023) (i.e., change in e-cigarette behaviors among the same individuals over time) (hereafter referred to as the “cohort study”).

Data for the panel study were restricted in order to assess the impact of the state law restricting the sales of flavored tobacco products. The “pre-law” data were from 11^th graders (Class of 2024) collected in fall 2022 (September 20, 2022–December 15, 2022; N=2686); the “post-law” data were from 11^th graders (Class of 2025) collected in fall 2023 (September 21, 2023–January 31, 2024; N=1027). To optimize comparability across the pre-law and post-law samples, we restricted the pre-law sample to participants attending the same 4 schools as students in the post-law sample (N=1215). Participants were excluded if they were missing data on past 6-month e-cigarette use (overall pre-law [n=5], analytic sample [n=2681]; restricted pre-law [n=3], analytic sample [n=1212]; post-law [n=1], analytic sample [n=1026]).

In the cohort study, data to examine changes in e-cigarette use patterns over time were restricted to any individuals who completed the fall 2022 data collection (i.e., Class of 2024 or Class of 2025; 10^th or 11^th graders) and reported past 30-day e-cigarette use at this wave (N=142), and who subsequently also completed the fall 2023 wave of data collection, approximately one year after the law. In sensitivity analyses, we examined the prevalence of past 30-day e-cigarette use and past 30-day use of any nicotine product at multiple waves (‘T’) (spring 2022, fall 2022, spring 2023, fall 2023, spring 2024) among those who reported past 30-day use of e-cigarettes at the prior wave (i.e., 6 months earlier; T-1) or two waves prior (i.e., one year earlier; T-2). These waves were selected to allow for incorporation of at least two sequential study waves (i.e., T-1 and T) for the combined sample including both the Class of 2024 and Class of 2025.

Ethics Statement

The study was approved by the University of Southern California Institutional Review Board (HS-19-00682). Written parental consent and student assent were obtained prior to data collection.

Measures

Pre- and post-law measures:

All participants were asked whether they had used e-cigarettes in the past 6 months (yes/no), and the number of days they had used e-cigarettes in the past month. We additionally assessed past 30-day cigarette use (yes/no), and past 30-day use of any other nicotine product (IQOS or heated tobacco products, smokeless tobacco, nicotine pouches, non-medicinal nicotine gums or lozenges, cigars, little cigars or cigarillos, or hookah; yes/no). We classified past 30-day nicotine product use as follows: no past 30-day use, use of e-cigarettes only, use of another nicotine product only, use of e-cigarettes and another nicotine product in the past 30 days. In sensitivity analyses, we also examined past 30-day use of any nicotine product (i.e., e-cigarettes, cigarettes, or any other nicotine product) (yes/no).

Participants were also queried as to how easy they felt it would be for their peers to purchase “tobacco-flavored e-cigarettes,” “mint or menthol-flavored e-cigarettes,” or “other flavored e-cigarettes” (not defined further) (assessed separately for a total of 9 items): (a) in person from a store, (b) online, or (c) from another person; participants rated each item on a visual analogue scale (VAS) from 0 (very easy) to 100 (very difficult). Response options of 50 (i.e., no movement of the indicator along the VAS indicating relative non-response) were removed from analysis.

Participants who initiated e-cigarette use in the 6 months prior to a given wave were asked the type of e-cigarette flavor that they used the first time they used an e-cigarette (flavorless, tobacco-flavored, menthol, mint, fruit, non-sweet, fruit-ice, candy/dessert, or another flavor). We assessed each flavor separately and dichotomized flavor to assess use of unflavored (flavorless or tobacco-flavored) vs. flavored e-cigarettes (menthol, mint, fruit, non-sweet, fruit-ice, candy/dessert) at first use.

Students who reported having used e-cigarettes in the past 30 days were asked about the source of the product, difficulty getting the flavors that they like to use (very easy, somewhat easy, somewhat difficult, very difficult; additionally dichotomized as very/somewhat easy vs. very/somewhat difficult), and flavor used most often in the past 30 days (among the same list of flavors as above, assessed separately and dichotomized as flavored vs. unflavored e-cigarette use).

Pre-flavor law measures (fall 2022):

Prior to the passage of the California flavor law, we queried participants on their overall support of the flavor law (strongly support, support, indifferent, oppose, strongly oppose) and their anticipated impact of the law on e-cigarette use amongst their peers (reduce use a lot, reduce use a little, no change in use, increase use a little, increase use a lot). Among those who reported use of e-cigarettes in the past 30 days, we also asked their perceived likelihood of doing each of the following if the flavor law were to be enacted, on a VAS from 0 (not at all likely) to 100 (very likely): using tobacco-flavored e-cigarettes, continuing to vape nicotine, switching to a different nicotine product.

Post-flavor law measures (fall 2023):

In the fall 2023 survey (approximately 9-12 months after the law was enacted), we asked participants about their perceived impact of the law on peer use of e-cigarettes (reduced use a lot, reduced use a little, no change in use, increased use a little, increased use a lot), and methods that their peers used to obtain flavored e-cigarette products (purchase nicotine and flavoring separately, purchase at an in-person store, use a delivery service, buy online, use another product, or “I don’t know anyone who has used flavored nicotine products”; select all that apply).

Covariates:

Data were collected on race/ethnicity, gender identity, and sexual identity (see Table 1).

Table 1.

Demographic characteristics of the analytic sample for the panel study, including those in 11^th grade at either the pre-law wave (fall 2022) or the post-law wave (fall 2023)

	Pre-Flavor Law Fall 2022		Post Flavor Law Fall 2023
	Overalla	Restrictedb	Restrictedc
	N (col%)	N (col%)	N (col%)
	N=2681	N=1212	N=1026
Race/Ethnicity
Asian/PI	755 (28.2)	505 (41.7)	463 (45.1)
Black/AA	34 (1.3)	15 (1.2)	13 (1.3)
White	193 (7.2)	82 (6.8)	63 (6.1)
Multiracial	186 (6.9)	108 (8.9)	64 (6.2)
Hispanic	1468 (54.8)	475 (39.2)	350 (34.1)
Another Race	34 (1.3)	25 (2.1)	9 (0.9)
Missing	11 (0.4)	2 (0.2)	64 (6.2)
Gender Identity
Male or Masculine	1102 (41.1)	516 (42.6)	429 (41.8)
Female or Feminine	1230 (45.9)	554 (45.7)	474 (46.2)
Transgender	22 (0.8)	11 (0.9)	5 (0.5)
Gender Variant/Nonbinary	88 (3.3)	35 (2.9)	30 (2.9)
Another Identity	22 (0.8)	15 (1.2)	12 (1.2)
Prefer not to disclose or missing	217 (8.1)	81 (6.7)	76 (7.4)
Sexual Identity
Heterosexual	1676 (62.5)	782 (64.5)	681 (66.4)
Bisexual or Pansexuald	448 (16.7)	203 (16.8)	145 (14.1)
Gay or Lesbian	65 (2.4)	27 (2.2)	26 (2.5)
Asexual	59 (2.2)	27 (2.2)	29 (2.8)
Questioning	109 (4.1)	50 (4.1)	37 (3.6)
Prefer not to disclose or missing	324 (12.1)	123 (10.2)	108 (10.5)

Abbreviations: N (col%) = Sample size and column percent

^aAmong participants in the Class of 2024 in 11^th grade, at all 11 schools

^bRestricted to participants in the Class of 2024 in 11^th grade, in 4 schools common to both samples in the panel study

^cRestricted to participants in the Class of 2025 in 11^th grade, in 4 schools common to both samples in the panel study

^dIncludes bisexual, pansexual, queer, demisexual, omnisexual, or another identity

Statistical Analysis

Descriptive analyses were used to present data from panel study, including findings from the pre-law wave (fall 2022) and the post-law wave (fall 2023, approximately 9-12 months after the law was implemented) among students who were in 11^th grade at the time of each data collection. Risk differences, 95% confidence intervals, and p-values are reported from adjusted regression models for binary outcome variables (past 30-day e-cigarette use, purchase location, and difficulty getting flavors [dichotomized]) in order to estimate the difference in risk at the post-law wave (fall 2023) vs. the pre-law wave (fall 2022) in the panel study. Betas, 95% confidence intervals, and p-values are reported from adjusted linear regression models for continuous outcomes (each perceived ease of purchase variable) to estimate the mean difference in outcome score for the post-law wave vs. the pre-law wave. All models are adjusted for race/ethnicity (Asian/PI, Hispanic, Another race), gender identity (male, female, another gender identity), and sexual identity (heterosexual, another identity) (categories collapsed due to small cell counts); missing data for covariates were handled using a missing indicator (i.e., separate category for missing data). Post-hoc power calculations were estimated for purchase location due to the small, restricted sample size for these analyses; estimates were based on observed effect estimates, the probability of the outcome among the pre-law sample, and the proportion of the analytic sample that was assessed post-law for the given outcome variable. In the cohort study, we describe nicotine use behaviors one year post-law (fall 2023) among the sample of individuals who reported past 30-day e-cigarette use at the pre-law wave (fall 2022) (N=142). In sensitivity analyses of the cohort study, we additionally described nicotine use behaviors (prevalence of past 30-day e-cigarette use, and prevalence of past 30-day use of any nicotine product) at 5 sequential waves (T) (spring 2022, fall 2022, spring 2023, fall 2023, spring 2024) among those who reported past 30-day use of e-cigarettes at the prior wave (T-1) or two waves prior (T-2) in the combined sample.

RESULTS

Participant Characteristics

Participants in the study were demographically diverse, and the samples were relatively similar (Table 1).

Differences in Behaviors and Perceptions, Pre-Law vs. Post-Law (Panel Study)

The prevalence of past 30-day e-cigarette use was slightly lower among 11^th-grade students after the law was enacted (4.1%) compared to prior to the flavor law (5.8%), although the effect was not statistically significant (p=0.11) (Table 2). New initiation within the past 6 months was also similar; 4.4% initiated pre-law, and 3.2% initiated post-law (p=0.23). At both time points, 100% of new initiators reported that their first e-cigarette was flavored (Figure 1). Among those who reported past 30-day e-cigarette use, 89.8% of the pre-law sample and 90.9% of the post-law sample primarily used a flavored e-cigarette. Among both new initiators and those who used e-cigarettes in the past 30 days, fruit and fruit-ice flavors were the most used e-cigarette flavors.

Table 2.

E-cigarette use and purchasing behaviors and perceptions in the panel study, including those in 11^th grade at either the pre-law wave (fall 2022) or the post-law wave (fall 2023)

	Pre-Flavor Law Fall 2022				Post Flavor Law Fall 2023
	Overalla		Restrictedb		Restrictedc
	Nd	N (col%) or M (SD)	Nd	N (col%) or M (SD)	Nd	N (col%) or M (SD)	aRD (95% CI) or aβ (95% CI)e	P-valuee
Past 30-day E-cigarette Use, N(%)	2681	157 (5.9)	1212	70 (5.8)	1026	42 (4.1)	−1.43 (−3.21, 0.35)	0.11
E-cigarette Initiation, N(%)f	2106	84 (4.0)	998	44 (4.4)	847	27 (3.2)	−1.08 (−2.83, 0.67)	0.23
Purchase location, N(%)g h i	157		64		42
Someone shares them with me		83 (52.9)		43 (61.4)		24 (57.1)	−2.94 (−20.5, 14.6)	0.74
I take them from someone		19 (12.1)		8 (11.4)		8 (19.1)	9.76 (−4.22, 23.7)	0.16
I buy them from someone		47 (29.9)		20 (28.6)		15 (35.7)	10.9 (−7.18, 29.0)	0.22
Buy online		13 (8.3)		9 (12.9)		3 (7.1)	#	#
Buy from any in-person retailer		27 (17.2)		12 (17.1)		10 (23.8)	3.55 (−12.5, 19.6)	0.64
Vape only		10 (6.4)		6 (8.6)		6 (14.3)	4.96 (−7.45, 17.4)	0.42
Vape and Smoke		12 (7.6)		4 (5.7)		7 (16.7)	#	#
Smoke shop		15 (9.6)		7 (10.0)		7 (16.7)	4.25 (−8.92, 17.4)	0.51
Convenience		11 (7.0)		4 (5.7)		4 (9.5)	#	#
Gas Station		10 (6.4)		2 (2.9)		5 (11.9)	#	#
Grocery store		5 (3.2)		1 (1.4)		3 (7.1)	#	#
Pharmacy		3 (1.9)		1 (1.4)		2 (4.8)	#	#
Liquor Store		9 (5.7)		4 (5.7)		6 (14.3)	#	#
Other		10 (6.4)		2 (2.9)		1 (2.4)	#	#
Difficulty in getting flavors, N(%)g	157		70		42			0.76k
Very easy		57 (36.3)		26 (37.1)		15 (35.7)	--
Somewhat easy		49 (31.2)		21 (30.0)		11 (26.2)	--
Somewhat difficult		9 (5.7)		5 (7.1)		4 (9.5)	--
Very difficult		5 (3.2)		3 (4.3)		1 (2.4)	--
Missing		37 (23.6)		15 (21.4)		11 (26.2)	--
Dichotomized	120		55		31			0.47
Very/somewhat easy		106 (88.3)		47 (85.5)		26 (83.9)	--
Somewhat/very difficult		14 (11.7)		8 (14.5)		5 (16.1)	5.65 (−10.6, 21.9)
Ease of Purchase, M(SD)j
In person from a store
Tobacco flavor	2,122	52.3 (31.8)	976	51.0 (30.6)	737	52.4 (29.0)	1.11 (−1.73, 3.96)	0.44
Mint/menthol flavor	2,079	51.7 (32.4)	958	49.9 (31.1)	728	51.8 (29.4)	1.61 (−1.30, 4.53)	0.28
Other flavor	2,132	45.9 (34.0)	981	44.6 (32.4)	746	45.2 (30.4)	0.42 (−2.55, 3.40)	0.78
Online
Tobacco flavor	1,998	42.1 (31.7)	919	40.8 (30.1)	710	39.9 (29.3)	−0.98 (−3.89, 1.93)	0.51
Mint/menthol flavor	1,978	41.2 (32.0)	912	39.2 (30.2)	706	39.0 (29.4)	−0.21 (−3.14, 2.72)	0.89
Other flavor	2,022	37.3 (32.1)	932	35.2 (30.0)	715	35.2 (29.7)	0.02 (−2.88, 2.91)	0.99
From another person
Tobacco flavor	1,999	31.7 (31.2)	921	31.4 (29.8)	712	34.7 (30.0)	3.03 (0.14, 5.93)	0.04
Mint/menthol flavor	2,011	31.1 (30.9)	931	30.2 (29.2)	706	34.1 (29.7)	3.60 (0.75, 6.44)	0.01
Other flavor	2,063	27.2 (30.1)	949	26.2 (28.1)	734	30.1 (29.1)	3.66 (0.95, 6.37)	0.008

Abbreviations: M=mean; SD=standard deviation; N (col%) = Sample size and column percent; RD = Risk Difference; CI = Confidence Interval

^aAmong participants in the Class of 2024 in 11^th grade, at all 11 schools

^bRestricted to participants in the Class of 2024 in 11^th grade, in 4 schools common to both samples in the panel study

^cRestricted to participants in the Class of 2025 in 11^th grade, in 4 schools common to both samples in the panel study

^dAnalytic sample

^eRisk difference is the difference in risk for each binary outcome for the post-law wave (2023) vs. the pre-law wave (2022), adjusted for race/ethnicity (Asian/PI, Hispanic, Another race), gender identity (male, female, another gender identity), and sexual identity (heterosexual, another identity); betas are reported for continuous outcomes and represent the adjusted mean difference in outcome value for the post-law vs. pre-law wave

^fNew initiation among individuals who reported no history of e-cigarette use at the prior wave

^gRestricted to participants who reported past 30-day e-cigarette use

^hSelect all that apply; answer choices do not add to 100%

ⁱPost-hoc power for analyses of purchase location ranged from 5.7% to 20.2%

^jEase of purchase of e-cigarettes; participants who responded “50” on the VAS were removed

^kP-value is a global test of effects from an adjusted multinomial logistic regression model

^#Value suppressed due to small sample size (n≤5 in any cell)

Figure 1.

Data from the panel study, including (a) First flavor used among new initiators pre-flavor law (fall 2022; N=27) and post-flavor law (fall 2023; N=21), and (b) most common flavor used in among those who have used e-cigarettes in the past 30 days pre-flavor law (fall 2022; N=70) and post-flavor law (fall 2023; N=42)

Among youth who reported past 30-day use of e-cigarettes, the most common source of access was someone sharing them (pre-law: 61.4%, post-law: 57.1%), followed by purchasing from someone (not from a store) (pre-law: 28.6%, post-law: 35.7%); the proportion reporting access through either method was not significantly different pre-law vs. post-law (ps>0.05) (Table 2). Few youth endorsed other methods of access to e-cigarettes. Notably, post-hoc power calculations for analyses examining change in purchase location ranged from 5.7% to 20.2%.

Overall, most reported that it was easy or very easy to purchase e-cigarettes in flavors they wanted both before the law (85.5%) and after the law (83.9%), with no statistical difference observed (p=0.47) (Table 2). Among all participants, youth generally felt that it was easiest for peers their age to purchase e-cigarettes from another person (lower scores on the VAS), followed by online, and by in person from a store. In general, across all purchase locations, students perceived that other types of flavored e-cigarettes were the easiest to purchase, followed by mint or menthol, and by tobacco flavors.

Prior to passage of the law, most 11^th graders either strongly supported (24.5%) or supported (31.2%) the law (restricted sample; Table 3). Participants generally felt that the law would either reduce use a lot (30.9%) or reduce use a little (45.9%). After the law was in place, nearly half of adolescents (45.5%) felt that there was no change in the use of flavored e-cigarettes among their peers. About a third of the sample felt that it reduced use a little (36.3%) but only 9.3% felt that it reduced use a lot.

Table 3.

Attitudes and perceptions regarding the California flavor law among 11^th grade students in the panel study, including those in 11^th grade at either the pre-law wave (fall 2022) or the post-law wave (fall 2023)

	Pre-Flavor Law Fall 2022				Post Flavor Law Fall 2023
	Overalla		Restrictedb		Restrictedc
	Nd	N (col%) or M (SD)	Nd	N (col%) or M (SD)	Nd	N (col%) or M (SD)
Pre-flavor law
Support of law, N(%)	2587		1178
Strongly support		643 (24.9)		288 (24.5)
Support		739 (28.6)		368 (31.2)
Indifferent		906 (35.0)		394 (33.5)
Oppose		218 (8.4)		91 (7.7)
Strongly oppose		81 (3.1)		37 (3.1)
Anticipated impact of law, N(%)	2461		1137
Reduce use a lot		754 (29.4)		362 (30.9)
Reduce use a little		1124 (43.8)		538 (45.9)
No change in use		439 (17.1)		172 (14.7)
Increase use a little		155 (6.1)		68 (5.8)
Increase use a lot		92 (3.6)		32 (2.7)
Perceived likelihood of the following, M(SD)e
Use tobacco flavored e-cigarettes	174	23.3 (29.2)	81	17.9 (23.8)
Continue to vape nicotine	171	36.6 (35.5)	77	30.0 (33.3)
Switch to a different nicotine product	167	31.3 (32.3)	76	29.2 (31.6)
Post-flavor law
Perceived impact of the law, N(%)					935
Reduced use a lot						87 (9.3)
Reduced use a little						339 (36.3)
No change in use						424 (45.4)
Increased use a little						49 (5.2)
Increased use a lot						36 (3.9)
Methods peers use to get flavored products, N(%)f					1027
I don't know anyone who has used flavors						702 (68.4)
Purchase nicotine and flavoring separately						65 (6.3)
Purchase at an in-person store						107 (10.4)
Use a delivery service						101 (9.8)
Buy online						155 (15.1)
Use another product						27 (2.6)

Abbreviations: M=mean; SD=standard deviation; N (col%) = Sample size and column percent

^aAmong participants in the Class of 2024 in 11^th grade, at all 11 schools

^bRestricted to participants in the Class of 2024 in 11^th grade, in 4 schools common to both samples in the panel study

^cRestricted to participants in the Class of 2025 in 11^th grade, in 4 schools common to both samples in the panel study

^dAnalytic sample

^eRestricted to participants who reported past 30-day e-cigarette use

^fSelect all that apply; answer choices do not add to 100%

Changes to Nicotine Use Behaviors Following Implementation of the Law (Cohort Study)

Among those who reported past 30-day e-cigarette use at the wave prior to the law, one year later (post-law), 46.5% reported no past 30-day use of any nicotine product at the post-law follow-up, 26.1% reported e-cigarette use only, 6.3% reported only using a nicotine product other than e-cigarettes, and 21.1% reported using both e-cigarettes and another nicotine product in the past 30 days at follow-up (Table 4). Post-law, most participants who had used e-cigarettes in the past 30 days reported using flavored e-cigarettes (93.3%) and also reported that it was easy or very easy to get e-cigarette flavors that they liked (90.0%).

Table 4.

Nicotine use behaviors one year post-law (fall 2023) in the cohort study, among individuals who reported past 30-day e-cigarette use at the pre-law wave (fall 2022)

Post-law (2023) Behaviors	Na	N (col%)
Past 30-day E-cigarette Use	142
No		75 (52.8)
Yes		67 (47.2)
Past 30-day Cigarette Use	142
No		116 (81.7)
Yes		26 (18.3)
Past 30-day Use, Combined	142
No nicotine use		66 (46.5)
E-cigarettes only		37 (26.1)
Another nicotine product onlyb		9 (6.3)
E-cigarettes + another nicotine productb		30 (21.1)
Most common e-cigarette flavor usedcde
Individual flavors	61
Flavorless		4 (6.6)
Mint		8 (13.1)
Fruit		19 (31.2)
Fruit-Ice		22 (36.1)
Candy/Dessert		7 (11.5)
Other		1 (1.6)
Dichotomized	60
Non-flavored		4 (6.7)
Flavored		56 (93.3)
Difficulty in getting flavorscf	60
Very easy		38 (63.3)
Somewhat easy		16 (26.7)
Somewhat difficult		5 (8.3)
Very difficult		1 (1.7)
Dichotomized	60
Very/somewhat easy		54 (90.0)
Somewhat/very difficult		6 (10.0)

Abbreviations: N (col%) = Sample size and column percent

^aAnalytic sample

^bIncludes cigarettes, IQOS or heated tobacco product, smokeless tobacco, nicotine pouches, non-medicinal nicotine gums or lozenges, cigars, little cigars or cigarillos, or hookah

^cRestricted to participants who reported past 30-day e-cigarette use

^dN=7 participants missing data on flavor used

^eNo participants reported use of tobacco, menthol, non-sweet, or another flavor

^fN=8 participants missing data on difficulty in getting flavors

Notably, the prevalence of both past 30-day e-cigarette use or use of any nicotine product in the past 30 days among those who reported past 30-day e-cigarette use 6 months or 1 year earlier did not appear to be affected by implementation of the law (Supplemental Table 2). For example, the proportion reporting no past 30-day nicotine use in fall 2022 (pre-law) was 44.4%, in spring 2023 (post-law, among those reporting use pre-law) was 43.8%, and in fall 2023 (post-law) was 44.7%.

DISCUSSION

The prevalence of past 30-day e-cigarette use (and new e-cigarette initiation) among 11^th-grade students in Southern California decreased slightly, though not significantly, after the restricted sales of flavored tobacco. This slight decline is consistent with national trends over the same time frame.²⁰ Critically, at both time points, 100% of students who had initiated e-cigarette use within the past six months reported doing so with a flavored e-cigarette, and about 90% of youth who reported past 30-day e-cigarette use primarily used flavored e-cigarettes. There were few changes in purchasing behaviors or perceived ease of access to flavored e-cigarettes before and after the law. Prior to the state law, students were largely in support of such a law and felt that it would reduce use among their peers. However, the post-law survey found that students generally did not believe the law had impacted use of flavored products. Among those who reported past 30-day use of e-cigarettes before the law, about half reported no past 30-day use of any nicotine product after the law, which is consistent with patterns of discontinuation among adolescents observed in periods prior to implementation, reflecting typical, intermittent adolescent use patterns rather than policy-driven cessation. The remainder generally continued to use e-cigarettes with or without other nicotine products, and very few youth reported switching completely from nicotine vapes to use of other nicotine products.

Overall, access to flavored e-cigarettes remained widespread, with little change in use of flavored products after the law was enacted. Our findings suggest that social sources play a crucial role in sustaining youth access to flavored e-cigarettes. Most adolescents reported that they obtained e-cigarettes from other individuals; purchasing directly from a retail location was less common. This finding underscores the challenge of restricting access solely through retail sales laws and highlights the need for complementary efforts targeting social distribution networks. Public education campaigns aimed at reducing social supply, along with increased penalties for individuals who distribute tobacco products to minors, may be warranted. While we did not collect detailed information on the chain of supply (e.g., individuals who provided access [family, friends, peers, etc.] or where social sources obtained flavored tobacco products), if the law had effectively reduced retail availability of flavored tobacco, we would expect to see a trickle-down effect on youth access through social sources (i.e., reduced flavored product availability and use). No such reduction was observed, suggesting potential gaps in enforcement.

The seemingly abundant and easy access to flavored e-cigarettes may also suggest challenges with implementation and enforcement efforts (which may impact both direct access in the retail environment, as well as access through social sources if such sources can similarly easily access flavored products). A study examining compliance with the sales prohibition found that 50.3% of vape shops at which store observations were conducted between April and July 2023 had flavored e-cigarettes available for sale.²¹ California has more than 22,000 tobacco retailers,²² and state compliance checks used to enforce laws on the sale of flavored tobacco products (and on the sale of tobacco products to minors) happen relatively infrequently at a given store.²³ In fiscal year 24 (October 2023-September 2024), the Department of Justice reported a total of 2,321 store visits in California (with just 2 visits during this time in the four cities in which schools in the current study were located).²³ However, this does not account for local enforcement efforts, for which data are not readily available. It may also take more time to achieve full compliance, and reduced access to and use of flavored nicotine products may be more evident over the next few years.

Our findings are consistent with prior research investigating the impact of prohibitions on the sale of flavored tobacco products on youth access and use. A cross-sectional study evaluating California’s law found no decrease in the prevalence of use among adolescents 12-17 years of age, with most participants reporting that it was ‘easy’ or ‘very easy’ to get flavors they like.¹⁸ A study of flavored tobacco sales restrictions in the San Francisco Bay Area (where many localities implemented flavor restrictions prior to the state law) found no decrease in the prevalence of e-cigarette use.²⁴ A study using national data to examine patterns of cigarette, cigar, and e-cigarette use among youth estimated no association of flavor restrictions on the prevalence of use of any nicotine product.²⁵ Researchers have also raised concerns regarding the potential for flavor prohibitions to result in a transition from e-cigarette use to use of other nicotine products – products which may be more harmful to health and result in a net decrease in the overall public health of young people.²⁶ We found that only 6% of youth who reported use of e-cigarettes in the past 30 days at the pre-law wave reported switching completely from e-cigarettes to another nicotine product post-law; it was far more common for youth to stop using e-cigarettes (46.5%), or continue using e-cigarettes either with (21.1%) or without (26.1%) other nicotine products. Notably, patterns of nicotine use among those who reported e-cigarette use at a prior wave did not appreciably differ when examining the prevalence of discontinuation of any nicotine product across multiple time periods, including prior to implementation of the law, from pre- to post-law, or entirely after the law was implemented, suggesting little impact of the law on discontinuation of nicotine use. The relatively high rate of discontinuation of nicotine use likely reflects typical use patterns among adolescents, who use nicotine relatively inconsistently, entering into and out of past 30-day use with relative frequency,²⁷ rather than cessation resulting from the law. Still, data on reasons for cessation or discontinuation were not collected, but would be informative in understanding the impact of sales restrictions on flavored nicotine products (e.g., some youth may have stopped using e-cigarettes because they were unable to access products in flavors that they like).

Research suggests that the success of tobacco control policies is heavily dependent on enforcement mechanisms, including compliance checks and penalties for violations.²⁸ Following the implementation of SB 793, California took additional steps to strengthen enforcement and close regulatory gaps. In October 2023, Assembly Bill (AB) 935 was signed into law, designating the California Department of Public Health as the primary enforcement agency and increasing penalties for non-compliant retailers.²⁹ The law broadens the definition of retail locations and includes vending machines and mobile points of sale. More recently, AB 3218 and Senate Bill (SB) 1230 were passed in September 2024, which further strengthened tobacco control in California.³⁰ Together, these new laws close additional loopholes (e.g., nicotine analogues such as metatine are included, the definition of ‘characterizing flavor’ now also includes ‘sensations’ [e.g, cooling sensation], restrictions on delivery sales), permit only the sale of FDA authorized tobacco products, allow seizure of prohibited (flavored) products, and impose greater penalties to retailers for the sales of flavored tobacco products and selling to those under 21 years of age.³¹ These legislative efforts underscore the state’s commitment to reducing youth access to flavored tobacco products through a multi-faceted regulatory approach. To further enhance the effectiveness of e-cigarette flavor laws, additional regulatory efforts could focus on expanding enforcement strategies to require more frequent compliance checks;additional research is needed.

Our study is subject to some limitations. The sample was restricted to 11^th-grade students from select schools in southern California, limiting generalizability to other regions across the state. Of note, the City of Los Angeles, unincorporated Los Angeles County, and several other local jurisdictions in southern California had passed legislation restricting the sale of some flavored tobacco products prior to the implementation of the state law. In the current study, no participants lived in locations where such laws were in effect prior to the state law. Differences in enforcement, social access, proximity to out-of-state retailers, or retailer compliance across regions could influence how effectively the flavor law reduces youth access to flavored products, which warrants further study. In addition, loopholes for online sales may have impacted access to flavored products after implementation of the state law. However, we found that few youth reported purchasing products online, and the proportion doing so did not increase after the law was in effect. Given that the participants were not legally able to purchase e-cigarettes, they may have been less directly impacted by the sales law, which is evident in our findings that most youth report that they access e-cigarettes through social sources. Finally, this study could not fully determine how youth continued to access flavored e-cigarettes post-law, as the sample of youth currently using e-cigarettes was small, cell sizes were insufficient to estimate changes in specific purchasing behaviors over time, and data on where social sources obtain flavored products were not collected. The estimated power for analyses of difference in prevalence of access by location ranged from 5.7% to 20.2%, providing further evidence that the study was underpowered to detect true differences between the two waves. Data on methods of access in a larger sample pre- and post-law would be helpful in clarifying whether access methods changed after implementation of the law. Future research should also examine how enforcement gaps along with secondary markets sustain youth access to flavored products.

CONCLUSIONS

The findings suggest that while California’s sales restriction represents an important step in reducing flavored tobacco use among youth, its effectiveness in reducing youth access to and use of flavored tobacco products (in particular, flavored e-cigarettes) is limited. In particular, 9-13 months after implementation, we observed little change in flavored e-cigarette use and report of easy access to these products. Notably, there appeared to be a very low likelihood of those using flavored e-cigarettes switching to other nicotine products after the law was enacted; instead, it was far more common for adolescents to report discontinuation of nicotine product use, a finding which appears to reflect typical adolescent use patterns rather than a direct effect of the law. Findings may indicate that stronger regulatory oversight and enforcement efforts at the point of sale are needed (i.e., changes to improve implementation of the policy), though more data are needed. In addition, targeted interventions to address social sources may serve to reduce youth access to and use of flavored e-cigarettes and other flavored tobacco products, particularly if paired with school and community education efforts.

What is Already Known on this Topic:

Flavored e-cigarettes are appealing to adolescents, driving initiation and sustained use. Limited data exist on whether policies to prohibit the sale of flavored tobacco products are effective in reducing adolescent access to and use of these products.

What this Study Adds:

Adolescent access to and use of flavored e-cigarettes remained virtually unchanged after implementation of a comprehensive law prohibiting the sale of flavored tobacco products in California.

How this Study Might Affect Research, Practice, or Policy:

In addition to comprehensive policies prohibiting the sale of flavored tobacco products, stronger regulatory oversight and enforcement efforts at the retailer point of sale, in addition to targeted interventions to address social sources, are needed to reduce youth access to and use of flavored e-cigarettes and other flavored tobacco products.

Abbreviations:

OR

odds ratio

CI

confidence interval