How is substance use portrayed in digital media and what impacts can it have

Bibliographic Data

ID21597500
AuthorsBenjamin C Riordan (0000-0002-7805-0652, Centre of Alcohol Policy Research La Trobe University Melbourne Australia, corresponding author), Damian Scarf (0000-0002-9486-8059, Department of Psychology University of Otago Dunedin New Zealand), Jessica E Merrill (0000-0002-8310-140X, Department of Behavioral and Social Sciences, Center for Alcohol and Addiction Studies Brown University School of Public Health Providence Rhode Island USA), Megan S C Lim (0000-0003-3136-6761, Burnet Institute Melbourne Australia), Emmanuel Kuntsche (0000-0001-8931-0790, Centre of Alcohol Policy Research La Trobe University Melbourne Australia)
Year2024
Volume43
Issue1
Pages5-8
Publication date2024-01-01
Open AccessYes
TypeARTICLE
VenueDrug and Alcohol Review (JOURNAL)
Journal identifiersISSN: 0959-5236 • E-ISSN: 1465-3362
PublisherWiley (PUBLISHER • GB)
DOI10.1111/dar.13784
PMID38017698
OpenAlexW4389149844
LanguageEN
References cited21

In 2021, Facebook Inc. changed its name to Meta. At first glance, the updated name simply reflected that the company had grown to own not just the Facebook platform, but also Instagram, Threads, WhatsApp and a desire to branch out and include virtual reality [1]. The term “meta,” however, has a deeper meaning. Meta is short for metaverse, a word coined by Neal Stephenson in his dystopian novel Snow Crash. The metaverse is a hypothetical future state where people “live” in a virtual world. It is a common trope in film and literature, including popular films like The Matrix and Ready Player One. Although a virtual-reality-based future appears unlikely, people do spend a considerable portion of their wakeful hours consuming digital media (e.g., social media, watching films, listening to music) [2]. One might wonder what people spend all their time looking at and whether it matters. In the context of substance use, the 12 papers in this special section aim to address both of these important questions. A recent meta-analysis found that posts that referenced alcohol (91.6%), poly-substance use (87.9%), e-cigarettes (85.9%), cannabis (79.5%) or tobacco (59.8%) were predominantly positive in nature [3], meaning that they primarily referenced “good” things about substance use. Only posts referencing opiates were more often negative (55.5%) than positive. Identifying how substances are portrayed on social media may be particularly important for understanding the link between social media and substance use, as several theories (e.g., social norms theory, social learning theory) suggest that greater exposure to substance-related posts (particularly those that are positive in nature) may contribute to the perception that substance use is common and acceptable, and/or that substance use leads to positive outcomes. In this special section, a range of research methods (e.g., focus groups, social media data collection), populations (e.g., adolescents) and social media platforms (e.g., Twitter, TikTok) were studied to dig deeper into how “positive” posts about substances may be. For example, Corcoran et al.'s [4] focus group study aligned with prior work and found that adolescents reported that alcohol-related posts on social media convey drinking as acceptable, normal and cool [4]. Similarly, Lim et al.'s [5] content analysis of medicinal-cannabis-related tweets showed that tweets touting the therapeutic value of cannabis are twice as common as those referencing intoxication and withdrawal. Both studies extend prior work by providing more nuance to the type of content that may be perceived positively by those who view it. Interestingly, both Whelan et al. [6] and Mathieson et al. [7] demonstrate that this tendency for substance use posts to be positive may not extend to all substances. Whelan et al. found that the plurality of MDMA-related TikTok videos were neutral (40.6%), but posts were still more likely to be positive (34.9%) than negative (17.9%) [6]. Likewise, Mathieson et al. report that gabapentinoid-related tweets tend to be neutral (48.4% gabapentin/68.2% pregabalin), but that negative tweets (30.0% gabapentin/22.7% pregabalin) are more common than positive tweets (21.6% gabapentin/9.1% pregabalin) [7]. Thus, while the papers in the special section reiterate the finding that the portrayal of popular substances (i.e., alcohol or cannabis) is predominantly positive, they also suggest that this may not be the case for all substances. Beyond manual content analysis of social media posts, emerging machine learning techniques allow researchers to analyse posts at scale [8]. Although sacrificing some degree of depth and nuance relative to deeper thematic analyses, machine learning can analyse extremely large datasets in close to real time [8]. In their systematic review, Fuller et al. [9] found that posts advertising the sale of illicit substances were extremely common on social media platforms, but they note that moving forward the challenge is to develop algorithms that can go beyond basic keywords and sentiment analysis to detect the increasingly sophisticated strategies that are used to advertise drugs on social media. Addressing this challenge will require collaboration between social and computer scientists to develop machine learning algorithms trained on more diverse and complex data (e.g., multiple substances, different countries, posts from multiple sites). Reflecting the wider literature [10], four papers in the special section provide an affirmative “yes” for the link between self-reported exposure to substance use on social media and substance use. Speaking to the robustness of the effect, the association between exposure to substance use holds across the various approaches, samples and measures of exposure [11-14]. While most previous research has focused on exposure to peers' substance use posting or substance use posting in general, three of the included studies focused on exposure to influencer alcohol posts. Influencers are often held in high regard by people who follow them and may be particularly influential as people emulate their behaviour and aspire to be more like them [14]. To measure the impact of influencer exposure, Strowger et al. [12] asked participants to list the top five influencers they followed and estimate how often they post about alcohol. They found people who reported greater alcohol exposure from influencers consumed more alcohol (quantity, peak alcohol use) even when controlling for alcohol exposure from close peers. Similarly, Steers et al. [11] reported positive bivariate correlations between frequency of liking or commenting on influencers' alcohol-related posts and peak drinks and alcohol-related consequences. Regarding mechanisms, Cristello et al. [14] provided some evidence that such associations may be explained by social norms, finding that the link between exposure to influencer alcohol posts and alcohol use was statistically mediated by injunctive norms (a similar mechanism noted for exposure to peer's substance use). One noted limitation with the current papers and the field at large is the reliance on self-reported exposure to social media content, which may be difficult for participants to remember and accurately report. Although some scholars have argued that perceived exposure may be more important than actual exposure [15], more accurate measures of exposure may be more convincing for prevention and intervention (e.g., to demonstrate the need to reduce exposure in order to reduce alcohol use). Therefore, future research should aim to provide more accurate evidence for the link between digital content exposure and substance use using objective measures of exposure to substance use content. Recent data collection approaches employing screen recording or intermittent screen shots may be a key next step [16, 17]. Similarly, experimental studies may offer additional evidence for the link between exposure and use by manipulating content people are exposed to in controlled settings (similar to Willoughby et al. [18] in the current special section). Given the link between exposure and use, reducing exposure to substance depictions in popular media may be a key strategy for reducing use. However, policy recommendations often focus specifically on reducing traditional advertising [19], neglecting the user-generated content or non-advertising content. For example, 86% of popular films included at least one depiction of alcohol and alcohol is even common in films aimed at younger audiences [20]. Patsouras et al. suggested that one barrier to implementing policies that aim to reduce or mitigate the impact of exposure is that people drastically underestimate how much alcohol there is in popular films [21, 22]. Thus, they aimed to determine whether informing people about the amount of alcohol in popular films and the impact of exposure would make people more supportive of restrictive alcohol-related film policies (e.g., including messages before a film about alcohol sponsorship or amount of alcohol depicted [22]). Although they found that participants who received the information did not rate the policies more favourably than participants in a control group, both groups were relatively supportive of several policies, particularly those focused on reducing exposure in films targeting youth (e.g., alcohol should be included in film classification/rating system). This suggests that the public is already supportive of the introduction of restrictive film policies to mitigate the potential harm of exposure. Several papers included in this special section identified that social media may offer a key platform for potential outreach and that monitoring how substance use is discussed may help identify areas that social media campaigns should focus on. For example, both Matheson et al. [7] and Whelan et al. [6] identified that the lack of harm-related posts indicated the opportunity for outreach to increase awareness about harms and Lim et al. [5] suggested that monitoring posts about medicinal cannabis could be important to counter unsubstantiated health claims. Fuller et al. [9] also noted in their systematic review that social media prevention campaigns were one of the most common interventions recommended in the papers they reviewed, but that thorough evaluations of these types of campaigns are limited. The evidence shows that these interventions may be effective, but they struggle to achieve the reach needed to create meaningful impact in an overcrowded market [23, 24]. Thus, although this approach is often recommended, it is a key area for future research to better understand how to effectively implement outreach campaigns and measure the effectiveness of these approaches. The included contributions highlight how substance use is discussed in digital media, the impact of exposure, and key areas for policy and intervention. They also identify key areas for future research, which should focus on implementing new methods like machine learning to better understand how substances are discussed in social media, more objective measures of substance use exposure, and additional work evaluating the effectiveness of public outreach campaigns on social media. An ongoing issue will be whether research can keep up with the metaverse (e.g., the development of new platforms [1]) and the added difficulty that a substantial amount of content is user-produced rather than paid advertisements. In the film The Matrix, people were tricked into living in a virtual world, completely unaware that their physical bodies were being used as an energy source. In the context of substance use, it will be increasingly important for people to have a critical lens on their digital media use, lest they be tricked into consuming and creating substance-related content glorifying substance use that advertisers could never get away with producing. BR wrote the first draft of the manuscript. DS, JM, ML, and EK provided feedback and revised the first draft. The authors declare no conflicts of interest

Aesthetics · Art · Dystopia · Internet privacy · Literature · Media studies · Metaverse · Social media · Sociology · Virtual reality · Wonder · World Wide Web · Behavioral Health and Interventions · Computer Science · History · Impact of Technology on Adolescents · Psychology · Social Psychology · Substance Abuse Treatment and Outcomes

Tools

Open DOIOpen Access