Adolescents’ perceptions of substance use messaging in the age of social media: resolving cognitive dissonance

Abstract

This study interviewed adolescents about their exposure to and perceptions of substance-related social media content and substance use prevention messages. Participants (analytic sample N = 30, age 14–18 years, in CA, USA, 40% male) were recruited from Instagram and Facebook for online semi-structured interviews. An interview transcript coding guide was developed based on the interview questions and emerging themes. Most (27/30) participants reported exposure to peers using substances on social media through posts made on personal accounts. All peer posts portrayed substance use in a positive light. Most participants reported exposure to formal prevention messages on social media (i.e. public service announcements) (19/30) and in schools (i.e. drug education) (21/30; 70.0%) teaching the negative consequences of substance use. Responses to the differences between peer posts and prevention messages included dismissing prevention messages (7/30), believing that their peers were more credible (4/30), desiring comprehensive substance information (3/30) and believing that the no-use message was ineffective for at-risk youth (4/30). Messages shared by peers online significantly contrasted with prevention messages (i.e. public service announcements and drug education). This difference appeared to undermine prevention message credibility. Balanced prevention messages acknowledging the spectrum of risk and reward when using different substances may reduce dissonance and increase engagement.

Introduction

The use of alcohol, cannabis and nicotine is common among high school students in the United States [1]. According to the nationally representative 2022 Monitoring the Future survey, by the time adolescents graduate high school, many 12th graders will have tried alcohol (61.6%), cannabis (38.3%) and nicotine vaping products (38.8%) [1]. Although the prevalence of alcohol use has been gradually declining, 13.6% of 10th and 28.4% of 12th graders have used alcohol in the past month and 12.6% have engaged in binge drinking (5+ drinks in a row) [2]. During the Coronavirus Disease 2019 (COVID-19) pandemic, binge-drinking rates decreased only slightly to 13%, challenging the idea that reducing adolescent substance use can be achieved solely by limiting sources of supply, such as substance-using peers and adults outside the home [2].

Early initiation of substance use increases risk of addiction (i.e. continued use despite negative consequences), morbidity and mortality [1, 3, 4]. These findings are troublesome and warrant increased attention to early prevention and education efforts to reduce risks and prevalence of adolescent substance use.

Nearly every school-based drug education program in the United States operates on the principle of universal abstinence, meaning that the information provided focuses solely on the adverse health, social and legal consequences of substance use to prevent initiation and promote total cessation [5–8]. There is substantial evidence for the rejection of abstinence-only drug education in schools and its limited effectiveness in changing teen drug-using behavior [9–17].

Cognitive dissonance, an uncomfortable state of tension caused by holding inconsistent cognitions [18], may be one reason why so many teens reject adults’ messages to abstain [1]. Abstinence-only educational strategies often fail to resonate with young people, partly because of the disconnect between the message and young people’s social context. Young people observe seemingly positive substance use experiences (e.g. seeing peers have fun at a party), which abstinence-only messages often do not address [19]. This contradiction between beliefs formed through peer-to-peer information and messages received from adults may contribute to the formation of cognitive dissonance. Consequently, youth perspectives on substance use prevention programs demonstrate a lack of trust in formal sources of information, with greater trust in information from peers [20].

In order to establish credibility and engagement, some experts propose acknowledging the positive aspects of substance use, that teenage substance use does not equate to substance use disorder, and that most teens do not become addicted [7, 21]. Known as a harm reduction approach, it presents abstinence as the safest choice but allows for flexibility in the provision of knowledge and skills to reduce risks of substance use. This acknowledgment of positive aspects of substance use is especially crucial in the digital age, where negative consequences are under-communicated by peers on social media [22].

Social media presents an important, new social context for teens and an outlet for prevention messages. American teens spend an average of >7 h per day on screen media for entertainment, more time than they spend in schools [23]. Smartphone ownership has increased as an inescapable element of teen life: 95% of teens now report that they have a smartphone or access to one and 45% of youth report being online on a near-constant basis [24]. Young people have historically been exposed to substance advertisements through media channels and brands that are attempting to sell products [25, 26]. While regulations for substance use advertisements have strengthened over the years to limit vulnerable populations’ exposure, today’s teens are increasingly being exposed to pictures and videos of substance use through friends’ social media accounts and other personal (i.e. non-commercial) social media accounts [18, 19]. One study found that 34% of the 10th grade students who were surveyed had at least one friend who talked about drinking online and 20% reported that their friends posted party/drinking pictures online [27]. Adolescents in the study reported on average that their exposure to this type of content was monthly or more frequent.

Because the nature of social media is performative, it is typically the positive aspects of substance use being portrayed, such as social belonging and identity formation [19, 27]. Peer-to-peer substance use prevention campaigns, such as Living the Example and JUULers Against JUUL [28, 29], portray the negative effects of substance use on social media using personal stories. Peer-to-peer messaging may decrease adolescents’ substance use intentions [29]. However, exposure to posts about positive experiences with substance use may be more frequent. For example, in an exploratory study of Instagram posts with hashtags #ecig or #vape, almost all of the 84 posts analyzed were promoting e-cigarette use, while none were critical of usage [19]. Almost half of these posts were generated from personal, rather than commercial, accounts.

Especially for adolescents, exposure to peers’ substance use increases interest in substance use and has been linked to future alcohol, cannabis, nicotine and tobacco use patterns [18, 27, 30–32]. In a study of college students, exposure to pro-alcohol and pro-cannabis social media content was associated with greater likelihood of current heavy episodic drinking and current cannabis use [18].

Messages shared through universal abstinence-only drug education have been criticized for contrasting with adolescents’ personal observations of peer substance use. This contrast can lead to cognitive dissonance and subsequent rejection and reduced credibility of prevention messages. Research has not considered how exposure to substance-related content on social media interacts with adolescent perceptions of prevention messages. A better understanding of young people’s exposure to, and perceptions of, substance use content online is necessary, as well as the degree to which they experience cognitive dissonance. Such information is vital for curating future prevention information that can be distributed both in-person and online and that remains credible across diverse realities of young people.

The present study interviewed youth who use social media to understand their exposure to substance-related content on social media and how this corresponds to their perception of prevention messages. Adolescent participants in CA, USA, engaged in semi-structured interviews about how they perceived and resolved discrepancies in messages about substances, with particular attention given to their observations of messaging on social media. We hypothesized that presentations of substance use from peers on social media would contrast with prevention messages from adults, causing cognitive dissonance and reducing receptivity to prevention messages.

Methods

Participants and procedure

Participants (analytic sample N = 30, described in Table I) were recruited through a Facebook and Instagram advertising campaign administered in October to November 2020. Facebook Ads Manager allows researchers to create advertisements targeting specific demographics based on age, location, interests, behaviors and more. In this study, we targeted users who were between the ages of 13 and 18 years and living in CA, USA. We used two Institutional Review Board-approved images and placed them on Facebook and Instagram feeds with a clear call to action button. Clicking on an advertisement redirected potential participants to an anonymous eligibility screening survey. Eligible participants were adolescents aged 13–18 years, living in CA, USA, who reported using Facebook or Instagram at least 4 days a week. Underage participants were required to include the name and email address of a parent/guardian. Adult participants and parents of underage participants received a consent form via email; underage participants received an assent form via email. On the consent form, adult participants and parents selected an interview format (Zoom audio-recording or Facebook/Instagram private messages). Facebook and Instagram were chosen for their private messaging features, which easily facilitated dialogue between the interviewer and participant. Interviews were ∼30 min long. All procedures were approved by the Stanford Medicine Institutional Review Board.

Table I.

Participant characteristics¹

ID	Age	Gender	Race and ethnicity
1	17	Female	Asian
2	17	Male	Asian
3	16	Male	Asian
4	17	Male	Asian
5	17	Female	Hispanic
6	17	Female	Asian
7	18	Female	Non-Hispanic White
8	18	Male	Asian
9	18	Male	Unreported
10	18	Unknown	Asian
11	18	Male	Asian
12	18	Male	Asian
13	18	Female	Asian
14	18	Female	Multiple races/ethnicities
15	18	Female	Non-Hispanic White
16	16	Female	Asian
17	16	Female	Asian
18	15	Female	Asian
19	17	Male	Hispanic
20	17	Male	Asian
21	18	Female	Asian
22	14	Male	Asian
23	17	Male	Asian
24	18	Female	Asian
25	15	Male	Hispanic
26	17	Female	Hispanic
27	18	Female	Non-Hispanic White
28	18	Female	Asian
29	15	Female	Non-Hispanic White
30	15	Female	Asian

¹Participant gender was inadvertently excluded from the demographic questionnaire. We estimated each participant’s gender using the US Social Security Agency’s list of top 1000 baby names (by gender) for 2005 (i.e. the year that most 15-year-old participants would have been born). For names that were not on the top 1000 list (n = 11, 36.7% of analytic sample), we used a ‘Baby Name Guesser’ website that determines the gender typically associated with a name using Google’s database (https://www.gpeters.com/names/baby-names.php). Finally, we ran a Google search for the two remaining unidentified names (0.7% of analytic sample). The gender of one name remained unknown, and the participant’s gender is designated as ‘unknown’.

An initial semi-structured interview guide was drafted by E.A.V. to explore adolescent experiences with substance use content on social media and was reviewed by the study team. Although the parent study focused primarily on nicotine vaping, R.H. added interview questions for the present analysis regarding the messages adolescents receive about all substance use, online and offline (e.g. ‘Are the messages you’re getting about drugs and alcohol on social media different from the messages adults have told you about drugs and alcohol? How so?’). The study team consisted of psychologists, a pediatrician and a school-based drug educator, all of whom work with adolescents. The guide was subsequently revised following pilot interviews conducted via Zoom and Facebook Messenger with three adolescent volunteers (aged 15, 16 and 18 years) who had participated in a previous community-based intervention.

Analysis

A transcription service was used to transcribe audio interviews conducted on Zoom. For interviews conducted on Instagram or Facebook, text interviews were downloaded from the platform. All transcripts were deidentified and then read by the authors E.A.V. and R.H. The same semi-structured interview guide was used in all interviews; however, some participants gave more detailed answers than others and provided richer data regarding their thoughts, feelings and experiences. Transcripts were ranked by E.A.V. based on semantic richness, from Tier 1 (richest content) to Tier 3 (least rich content) [33].

A coding guide was developed in an iterative process to synthesize participants’ experiences. The initial coding guide was developed based on the semi-structured interview guide. E.A.V. drafted the initial coding guide in the form of a table that listed initial themes, with definitions and examples. The theme relevant to the present analysis was ‘messages about substance use’. The results of other themes, which are specific to nicotine vaping, are published elsewhere [34]. R.H. coded each transcript based on the initial coding guide and took notes on emergent subthemes. Content in codes was cross-checked with E.A.V., and new, emergent themes discovered in the data (e.g. message sources and congruence of messages received from adults versus on social media) resulted in a revised coding guide. Three Tier 1 transcripts were then coded by E.A.V. and R.H. independently using the revised coding guide. Content in subthemes was cross-checked by the authors, and the coding guide was further revised. The authors coded three additional transcripts at a time until no new subthemes emerged.

E.A.V. coded Tier 1 and Tier 2 transcripts. R.H. independently coded a randomly selected subset or Tier 1 and Tier 2 transcripts (n = 9, 30%). Due to high intercoder agreement (97.7%), we used the senior author’s (E.A.V.’s) codes for analyses. Determination of transcript richness was inherently subjective; therefore, all Tier 3 transcripts were reviewed by E.A.V. to ensure that saturation had been reached and no new themes or subthemes emerged. The analytic sample consists of participants with Tier 1 and Tier 2 transcripts (i.e. transcripts richest in content). Emergent subthemes relevant to message perception are described later. Some quotes were edited for brevity and clarity. To protect participants’ confidentiality, interview transcripts are not available.

Results

Source of messages

Participants described the messages they received about cannabis, nicotine and alcohol. Participants discussed prevention messages delivered from formal sources [i.e. drug education in schools and public service announcements (PSAs) delivered on social media] and messages posted on personal social media accounts (i.e. personal posts).

Formal sources

Participants were asked about the types of messages they had received in school-based drug education (‘drug education’) and social media prevention messages (‘PSAs’). Most participants (21/30; 70.0%) described being taught not to use substances in schools. A few participants (8/30; 26.6%) reported how they learned about adverse consequences, such as addiction, lung cancer or liver damage, in school-based drug education. One participant spoke about how all the messages in drug education promoted abstinence only.

A majority of participants (19/30; 63.3%) reported seeing prevention messages shared on social media (i.e. PSAs), showing the negative effects of substances. PSA campaigns were reportedly seen on YouTube (5/30; 16.6%), Instagram (6/30; 20.0%), Snapchat (3/30; 10.0%) and TikTok (2/30; 6.6%). In terms of frequency, Participant 22 reported seeing PSA social media advertisements on ‘a daily basis, maybe like two to three times a day’. Anti-vaping advertisements were the most commonly recalled PSAs by participants.

When asked to describe the messages of the PSAs, Participant 16 recalled:

I get a ton of ads on both Instagram and Snapchat informing against vaping, saying, ‘Oh, if you vape, you could be inhaling toxic metals that harm your lungs,’ or a testimony of some teenager saying something along the lines of, ‘Vaping was my big mistake; don’t do it,’ or something like that.

Participants reported that PSAs commonly sent messages disapproving of vaping as a coping mechanism, portraying teenagers vaping and losing self-control, and demonstrating the negative impact of vaping on relationships, mental health and physical health.

Interestingly, transient PSAs on ‘stories’ (i.e. social media content available to view for a limited time) would sometimes emerge adjacent to personal posts that promoted or idealized substance use. One participant (Participant 6) recalled how a ‘Don’t Vape’ message came up right after they saw a personal post on Instagram showing someone vaping at a party. The opportunity for strategic placement made the participant optimistic that prevention messages on social media were more influential than personal account posts that promote use:

I think even if you see someone smoking, the chances of you going and also smoking are quite less compared to if you see the effects or you see what harm it can do.

Personal accounts

Almost all participants (27/30; 90.0%) reported seeing images and videos of peers using alcohol, vaping products or cannabis posted on other teenagers’ social media accounts (i.e. ‘peer posts’). Most participants (26/30; 86.6%) described peer posts as portraying substance use in a positive light. When asked to describe the indirect messages peer posts convey about substance use, participants described portrayals of substance use as glamorized, mature, laidback, chill or carefree. Posts often depicted clouds of smoke or vapor, with music and lights in the background, usually in a social or party setting. The type of substance use shown in these posts was perceived as recreational and not problematic. Participants consistently reflected that they perceived teenagers in the posts as having a good time while vaping, drinking or smoking.

When asked if there was a message attached to these posts, Participant 4 mentioned that the posts are promotional and made a parallel to ‘rap videos’. Both types of posts would glamorize substance use by likening it to a happy mood but would not directly tell viewers what to do:

They’re not really promoting it because it’s illegal. But they‘re obviously showing they’re benefitting from using those substance. They don‘t explicitly like saying [sic] like I’m having so much fun with my friends and stuff. But the way they do it looks like it’s a necessity to be a part of a certain acts, like let’s say a party. Then, it’s like, look. Like this is how the party should look like.

Several participants worried that teens may form one-sided beliefs about substances because peers online do not talk about the negative health risks of drugs. Participant 28 described a consequence of viewing pro-substance content on social media:

Some teens may only see posts that portray drugs in a positive light, so their knowledge on the effects of drugs could be very limited. As a result, they only focus on the possibilities of feeling good and showing off.

If the posts do have a caption, Participant 19 reported that usually they just say something like ‘We’re vibing’. Participant 2, elaborated that some of these posts have captions such as ‘Wow, this was so fun’ or ‘All I need is my vape’.

Participant 8 described that peer posts with substances made teenagers seem more rebellious and they worried that this risk-taking perception ties into the idea that vaping is ‘a cool thing to do’ or ‘a way of breaking from the norm’. Posting substance-using content on personal accounts was also a way Participant 8 believed that teenagers ‘show who they are and the group that they fit into’.

Congruence of messages

When participants were asked whether messages received from peers on social media differ from those received from prevention sources, two themes emerged. Participants discussed congruencies and incongruencies between formal prevention messages (i.e. drug education and PSAs) and peer posts.

Congruent messages

Most participants (20/30; 66.6%) recalled that when formal prevention sources share messages about substances, they solely focused on the negative consequences. A number of participants (6/30; 20.0%) discussed the formal prevention sources (i.e. PSA and drug education) being consistent in their abstinence-only message.

When asked about this congruence, Participant 30 described how both PSA and drug education are ‘telling me to like not like do drugs and stuff’. These formal sources of messaging differed in language and content while conveying the same overarching abstinence message. Participant 8 shared that in their experience, school-based messages emphasize the adverse effects of using substances, while prevention messages on social media emphasize that substance use is uncool.

Incongruent messages

No participants described peer posts and formal prevention messages as being congruent with each other. Instead, participants (20/30; 66.6%) described incongruences between messages received from prevention messages and those that came from peer posts. Prevention messages were described by Participant 18 as ‘strikingly different from what I received from social media’ via peer posts. Several participants (7/30; 23.3%) discussed how they resolved the inconsistencies between prevention messages and personal posts by coming to the conclusion that prevention messages were not accurate, credible or helpful. Another minority (4/30; 13.3%) conferred more credibility to prevention posts and dismissed their peers. Several participants discussed wanting more complete information and safer use guidelines.

Positive versus negative messages.

Participant 24 described how the substance use they observed on peer posts did not reflect what they have learned from prevention messages:

We are taught from a young age that drugs and smoking are not cool and harmful to the mind and often lead to bad decision making. That is the opposite of what social media [peer posts] promotes.

Again, Participant 15 discussed the incongruence between prevention messages and those that came from teenagers’ personal accounts:

It’s sort of the same message but opposing behavior. A lot of the anti-vaping ads are like it’s cool to not vape. But then people’s Snapchat posts are like it’s cool to vape.

Participant 12 described how “fear-mongering” made prevention messages seem unrealistic and inapplicable to their experiences:

I feel there’s a little bit of a fear-mongering factor. It’s over-dramatized to the point where it becomes unrealistic. The message kind of flies over your head because it doesn’t apply to you because it’s like, ‘Oh, I would never be someone like this who would turn from marijuana to like cocaine or something’ and so on and so forth.

Early experimentation versus established use portrayed.

Another incongruence was the seriousness of substance use posted on personal accounts versus what was shared in prevention messages. This appeared to call into question participants’ perception of health risks. Participant 5 reflected on this incongruence:

I feel like even though they say that they bring a lot of health problems, I’ve never really seen anyone having any issues from it.

Participant 14 describes how they never see the warnings they have heard in prevention messaging such as ‘it’s going to ruin your brain’ manifesting in their friends’ substance use:

I see all my friends doing it and they’re not hospitalized. You see everyone smoking weed and they’re completely fine.

Because the substance use teenagers were seeing on personal accounts online focused on the positive aspects of use, Participant 8 worried that their peers would dismiss the prevention message by coming to the conclusion that the warnings were untrue:

When other people are doing it, how can it be something bad, the school is probably just wrong and it’s something that’s harmless.

Despite the incongruences, some participants (4/30; 13.3%) discussed trusting prevention messages. Participant 30 felt that the consistency between formal prevention messages about the dangers of vaping and addiction resulted in their aversion toward substance use:

My parents taught me from early on to not use [substances]. And then also at school my health teacher talked to us about the consequences, so I’ve decided to stay put and not use.

A few participants (3/30; 10.0%) described how they dismissed, discredited or unfollowed personal accounts that posted about substances.

Desiring more complete information.

A number of participants (4/30; 13.3%) believed that prevention messages did nothing for those who have already initiated use. They questioned the disconnect between goals of abstinence and teenagers’ need for safer use strategies that meet them where they are at. Participant 3 described the mismatch between prevention efforts focusing on abstinence, which sent messages that ‘drugs are bad’, and the reality that their friends were already using substances and were not likely to stop:

It’s two different worlds. My friends get all those messages, but they still do it, so is it really working?

Almost all participants (27/30; 90.0%) had been exposed to situations where substance use was occurring, both online and in-person. While they believed that prevention messages exaggerated risks, participants acknowledged that there were real consequences to substance use and worried that their peers who chose to engage were not prepared to navigate the ambiguities of each substance. Participant 17 named several scenarios their peers may encounter wherein they would need more complete information about substance effects and harm reduction strategies but were not provided this information from formal prevention sources:

They don’t really teach what’s going to happen if you already have done it. Let’s say, you have alcohol poisoning, then what do you do? Or someone’s having a really bad trip, then what do you do? The message that’s given off is don’t do it and you’ll never have to find out.

Participants believed that adults may be driving young people to discredit warnings and use substances recklessly when safer use and harm reduction messages are excluded from prevention efforts.

Discussion

In today’s digital environment, where young people are frequently exposed to positive portrayals of substance use, this study solicited adolescents’ perceptions of messages they receive about substance use online and offline. Interviews with adolescents in CA, USA, yielded insights into the differences in messages posted by formal prevention sources (i.e. school-based drug education and PSA) and by peers on personal social media accounts. Participants reported that peers shared overwhelmingly positive portrayals of substance use, while prevention sources only portrayed the negative consequences of substance use and urged abstinence. The discrepancy between the two sources seemed to undermine prevention credibility, perceived effectiveness and relevancy.

In addition to learning about the negative effects of substances in schools, almost all participants (19/30; 63.3%) reported seeing daily anti-substance PSA messages shared on social media. Participants (6/30; 20.0%) described consistency between school-based drug education and online PSAs in their negative health effect messaging and no-use goals. Several (4/30; 13.3%) described how prevention messages were more credible because of the consistency across what was taught at home, in school and now online through PSAs. While prevention sources focused solely on sending messages about the negative consequences of substance use, most participants were also exposed to social media content on personal accounts portraying peers using alcohol, vaping products or cannabis in a positive light. No participants described what they observed being posted by peers as being congruent with prevention messages. Consistent with prior research on drug education and cognitive dissonance, adolescents in this study discussed how they resolved the inconsistencies between prevention messages and observations of their peers by concluding that prevention messages were not credible or accurate [35].

Adolescents identified a mismatch in seriousness of risks, messages and strategies between abstinence-oriented prevention efforts and what teenagers observed in their social contexts. Almost all participants (27/30; 90.0%) had been exposed to peers using substances, online or in-person. Participants discussed how prevention messages warned about addiction, overdose and death, but these consequences were not observed by participants when they reflected on their peers’ substance use, online or in-person. Instead, participants discussed the substance-related consequences they could encounter, such as alcohol poisoning or difficult psychological experiences, but worried that they were not given any suggestions from prevention efforts on how to navigate these situations other than ‘just don’t do it and you won’t have to find out’. Consistent with behavioral science principles, participants in our study believed that prevention goals and messages must be tailored to the behavioral and developmental stages of the audience [16]. In other words, they needed prevention efforts to ‘meet them where they are at’—a key philosophy of a harm reduction approach.

Without information about safety and harm reduction, participants in this study warned that prevention messages shared through drug education and PSAs may be driving young people to discredit adults’ warnings and use substances recklessly. This was expected, as another study that surveyed youth about their perspectives on substance use prevention approaches found that teens prefer messages that encourage setting limits and practicing moderation, believing that nuanced messages are more realistic and effective for their contexts [20].

Implications

This study adds to the literature on youth perceptions of substance use messaging by exploring the stark contrast between prevention messaging and messages from portrayals of substance use from peers online. The reality is that by the time they are in secondary schools and reached by universal prevention programs, many teens will have already engaged in or observed some substance use. Receiving abstinence-only prevention messages that portray all substances as equally dangerous, despite the reality of a spectrum of risk, is inadequate and contradictory. Teens see and experience the costs and benefits of substance use. Stories of the terrible side effects of substances ring hollow alongside their own and others’ experience of substances. Consequently, teen perspectives on current substance use prevention programs demonstrate a lack of trust in formal sources of information, with higher trust in information from peers [20].

For drug prevention messages to be credible, they must be consistent with adolescent realities; otherwise, they may generate mistrust from adolescents, especially those who are more experienced and at-risk for substance use problems. Being consistent means acknowledging the positive effects of substance use [20] and that use does not always lead to addiction [1]. Messages could offer harm reduction strategies, such as practicing moderation by setting limits for those who have already initiated [36]. The literature on cognitive dissonance suggests that messages about moderation are more likely to be accepted [37]. Importantly, messages must be tailored to the unique properties and consequences of each substance and to the intended audience. Some substances, including some that are commonly used by youth, are not conducive to moderate, casual use. For example, nicotine vaping devices popular among youth (e.g. JUUL [38]) are highly addictive, especially to developing brains [39]. Youth who are using nicotine and other highly addictive substances should be supported in pursuing abstinence.

In school environments, harm reduction has shown to increase youth engagement and endorsement of content compared to abstinence-only messaging [21, 40, 41]. However, harm reduction approaches have yet to be applied to PSA messages delivered on social media, an opportunity that requires attention in future research and practice. The opportunity for social media to reach young people online, or meet them where they are at, is substantial. Against the backdrop of pro-substance messages posted from personal accounts on social media, PSAs presented adjacently can promote critical reflection on risks versus benefits and counter the development of lenient cognitions toward substance use, which have been found to escalate usage and encourage uptake in use [42–44]. Embracing a harm reduction framework in PSA message development broadens the goals, possibilities and potential for reaching and empathizing with teenagers’ behaviors.

Limitations and future directions

The findings of this study should be considered in light of its limitations. First, the results are limited to adolescent recall of messages and images. Second, this study examines perceptions of substance use–related content and messages but does not examine current behavior. These findings on exposure to congruent and incongruent messaging may not be generalizable to actual substance use. Third, participants were all adolescents in CA, USA. Most identified as Asian and reported minimal substance use. Therefore, the results may not generalize to all US teens. Lastly, the presence of an interviewer can bias results; however, we took steps to prevent this. The interviewer (R.H.) had extensive experience working with adolescents and creating a nonjudgmental environment for them to express their thoughts, feelings and behaviors regarding substance use. At the beginning of each interview, the interviewer emphasized that we value participants’ perspectives and any experiences they want to share are important to us.

Conclusion

This study suggests that youth exposure to positive portrayals of substance use on social media has exacerbated the gap between prevention messages and what teens believe and observe. The adolescents in this study appeared to discredit or disapprove of abstinence-only prevention efforts (drug education and PSAs) because the messages lack relevance to their personal observation and realities of their lives. Cognitive dissonance theory provides a possible explanation for the reduced credibility and rejection of messages that have been found in previous drug education research. To prevent the loss of credibility, it is important that prevention messages are relevant to young people’s beliefs, behaviors and realities. The expansion of prevention messages, goals and strategies, from abstinence-only to harm reduction, may be even more imperative for staying relevant in adolescents’ online and in-person realities given their increased exposure to pro-substance messaging on social media.

Conflict of interest statement

None declared.