Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 May 1.
Published in final edited form as: Addict Behav. 2024 Jan 17;152:107956. doi: 10.1016/j.addbeh.2024.107956

Associations between Frequency of Exposure to Peer-Generated Alcohol-related Posts and Alcohol Use within a Social Network of College Students

Megan Strowger 1, Matthew K Meisel 1, Michelle Haikalis 1, Michelle L Rogers 1, Nancy P Barnett 1
PMCID: PMC10988997  NIHMSID: NIHMS1963392  PMID: 38301589

Abstract

Peer alcohol use, commonly assessed via perceptions of how many drinks peers consume, is a robust predictor of college drinking. These perceptions are formed by in-person exposure to peer drinking but also may be affected by seeing alcohol-related content (ARC) shared on peer social media accounts. Most research assesses exposure by asking about the frequency of ARC sharing by a whole friend group, potentially missing influences from specific friends. Social network methods collect information about specific friends and their behavior but few studies have used these methods to examine the effects of ARC on drinking, nor have they examined potential moderators of this relationship. The purpose of this study was to examine whether perceived frequency of exposure to ARC shared by social network members on social media is associated with participant alcohol use after controlling for network members’ self-reported alcohol use, and if participant gender and relationship qualities with network members moderate this association. Participants were 994 college students (Mage = 21.17, SD = 0.47; 61.8% female; 55.4% White; 12.3% Hispanic) who completed a web-based survey. Due to the social network design, network autocorrelation analyses were conducted, which revealed that greater perceived frequency of exposure to network member ARC was significantly associated with higher alcohol quantity above and beyond network members’ alcohol use. Peer ARC had a unique association with drinking behavior independent of in-person peer alcohol use, although the cross-sectional design precludes making causal inferences. Clinicians delivering alcohol interventions to college students may wish to discuss exposure to ARC as another important source of peer influence and how media literacy may help reduce the effects.

Keywords: Alcohol, Alcohol-related content, College students, Social media, Social network

1. Introduction

Alcohol use is common among college students, with 56% reporting past 30-day use (Schulenberg et al., 2021). Additionally, heavy episodic drinking (i.e., consuming five or more drinks on an occasion) among college students remains pervasive, with 24% reporting binge drinking in the past month (Schulenberg et al., 2021). Alcohol use is associated with numerous negative consequences (e.g., missed classes, hangovers, driving under the influence; Conway & DiPlacido, 2015; Patrick et al., 2020). Given persistent high rates of hazardous alcohol use among college students, it is critical to examine factors that exacerbate alcohol use in this population so effective interventions can be developed to reduce alcohol-related risks. Peer alcohol use, commonly assessed via perceptions of how many drinks peers consume, is a robust predictor of college drinking (Borsari & Carey, 2001, 2003; Lac & Donaldson, 2018). These perceptions are in part formed by in-person exposure to peer drinking but also may be affected by seeing alcohol-related content shared on social media.

1. 1. Social Media and alcohol use

Alcohol-related content (ARC) is commonly found on college students’ social media profiles or accounts (Kiciman et al., 2018; Moreno et al., 2021) across a range of social media platforms (e.g., Twitter, Kiciman et al., 2018; Facebook, Moreno et al., 2021; Instagram, YouTube, Reddit; Cirillo et al., 2022), with students commonly reporting that they see ARC on more than one platform (Cirillo et al., 2022; Strowger et al., 2023; Ward et al., 2022). ARC typically refers to social drinking situations (Hendriks et al., 2018; Ward et al., 2022) and highlights the positive effects of alcohol use (e.g., having fun, stress relief). Both sharing and viewing ARC have been linked to greater alcohol use (for systematic reviews see Curtis et al., 2018; Gupta et al., 2016), with exposure to peer content having stronger associations with alcohol use than content shared by brands or influencers (Concoran et al., 2023; Roberson et al., 2018; Strowger et al., 2023).

1.2. Social network methods for understanding peer influences on drinking

Most of the research on exposure to ARC uses a descriptive norms approach in which participants are asked to report the frequency of ARC sharing by their friends (Strowger & Braitman, 2022). This type of global question allows researcher to broadly examine social influence but does not specify the friend group well. Specifically, it is unknown what individuals are considering when responding to a question about how often their friends post ARC.

Social network methods prompt participants to provide an answer about each person they identify, thus providing more granular information about individuals who are influential for alcohol use (for systematic reviews see Knox et al., 2019; Rinker et al., 2016). An egocentric social network approach collects participant reports/perceptions of specific friends, but self-reports from friends are not collected. In two studies using these methods, perceived exposure to ARC shared by important social network members was associated with greater alcohol consumption (Huang, Unger, et al., 2014; Strowger et al., 2022). A sociocentric social network approach entails collecting self-report from all members of a complete network and relationship ties within the network (Carolan, 2014) providing information than cannot be achieved using an egocentric approach; a sociocentric approach also allows for the collection and analysis of characteristics of the nominated peers and the nature of the ties to those peers. The only study to use sociocentric methods found that ARC exposure among high school students was not linked to alcohol use over time, although participants were more likely to later form friendships with those who had similar levels of ARC exposure (Huang, Soto, et al., 2014). In sum, the limitations of the recent work on the relationship between alcohol use and exposure to peer ARC include using global (i.e., whole group) items, and not considering how friendship characteristics and qualities may modify this association; both of these limitations can be addressed by using a social network approach.

1.3. Sex, relationship characteristics and alcohol use

Historically, gender differences in alcohol consumption are observed with males generally engaging in heavier drinking than females (SAMHSA, 2022). The sex of the ARC viewer also appears to moderate associations between ARC exposure and alcohol quantity among college students. However, the evidence is mixed with some studies showing greater exposure to ARC is more strongly related to alcohol use only among males (Boyle et al., 2016; Roberson et al., 2018) and others among both males and females but at different times of the academic year (Boyle et al., 2016; Davis et al., 2021). Specifically, Davis et al. (2021) found exposure was significantly associated with drinking among males pre-matriculation, but for females this association was only significant during the first semester of college. These prior studies examined global ARC exposure from peers broadly defined. Peer influences from proximal sources (e.g., important peer) impact college drinking more than distal sources (e.g., same-age peer; Borsari & Carey, 2003). These limited inconsistent findings underscore the need to further examine sex-specific effects of ARC exposure from important peers on college drinking.

Relationship characteristics such as closeness are potentially influential factors in the development of alcohol use and problems. Based on social learning theory (Bandura & Walters, 1977), Borsari and Carey (2006) proposed there were three components of high-quality peer relationships: stability (size of network), intimacy (closeness), and support (frequency of communication and socializing with peers). Borsari and Carey (2006) found these components moderated associations between social learning theory components (social reinforcement, modeling, cognitive processes, reciprocal determinism) and college drinking. In a sample of adolescents, closeness to peers who used substances predicted greater alcohol use (Cruz et al., 2012) but in another, closeness did not moderate the effects of peer substance use on personal alcohol use (Mason et al., 2017), possibly due to well-established friendship networks. There is also evidence that frequency of socializing with friends is associated with greater alcohol use (Finlay et al., 2012). Given that most ARC (67%) features in-person drinking situations (Hendriks et al., 2018), social learning theory may help explain the process that occurs when people view ARC including how these relationship qualities of closeness and frequency of communication may moderate the association between exposure to ARC and drinking behavior.

In addition, having more drinking buddies (i.e., friends one regularly does activities with centered around drinking) predicts alcohol use and problems, even after controlling for having network members who drink (e.g., Lau-Barraco et al., 2012; Reifman et al., 2006). Most importantly for the present study, having a higher proportion of drinking buddies who share ARC is associated with greater drinking frequency (Strowger et al., 2022). Taken together, these findings suggest that qualities of peer relationships modify associations between ARC exposure and alcohol use. However, these qualities and their associations have not before been studied in a large sample using sociocentric methods.

1.4. The current study

The current study had two aims: (1) to investigate whether perceived frequency of exposure to ARC shared by social network members on social media is associated with participant alcohol use after controlling for network members’ alcohol use and participant’s own sharing of ARC (RQ1), and (2) to examine whether participant gender and relationship qualities with network members (closeness, frequency of communication, frequency of drinking with) moderate the association between network member ARC sharing and participant alcohol use (RQ2).

2. Method

2.1. Participants

A total of 994 participants (Mage = 21.17, SD = 0.47) took part in the study. The sample comprised 370 (37.3%) men (cis- or trans-), 602 (60.6%) women (cis- or trans-), 17 (1.7%) non-binary/genderqueer/gender non-conforming individuals, and 4 (0.4%) with a different gender identity. Birth sex was 377 (38.2%) male and 610 (61.8%) female. Racial identities reported were White (n = 551, 55.4%), Asian (n = 239, 24.0%), Black (n = 73, 7.3%), Native American or Alaskan Native (n = 2, 0.2%); those who endorsed more than one racial identity were categorized as multiracial (n = 86, 8.7%) and some participants did not respond (n = 43; 4.3%). Hispanic ethnicity was reported by 122 (12.3%) participants. All participants were members of the class of 2021 who attended a mid-sized, private university in the northeastern United States.

2.2. Procedure

Data were obtained from a larger investigation examining social influence and alcohol consumption in a class year cohort. All students in the original class of 2021 were invited to participate in a longitudinal social network study, beginning in the fall of their third (i.e., junior) year1. Participation consisted of completing web-based surveys, administered at five timepoints from fall 2019 to fall 2021. Data for the current analyses were drawn from a single timepoint, collected in the fall 2020 semester when most students were in their senior year and on campus. At this wave, 1,744 students were eligible to participate2. Of those, 1,383 (79.3%) students were enrolled in the study, 1,265 completed the survey (91.5%)3. The sample was reduced due to missingness on key variables (see measures section) with the final N = 994. Participants received financial compensation of a $60 gift card at this wave. All procedures were approved by the Brown University Institutional Review Board.

2.3. Measures

2.3.1. Demographics

Participants provided information about age, gender identity, sex assigned at birth, race, and ethnicity. For race, participants could select all identities that applied to them. A multiracial category was created by the researchers to reflect participants who had selected more than one identity. The small sample size for those with non-binary or different gender identities (n = 17) precluded our ability to statistically evaluate gender differences with this distinct group. Rather than dropping these participants, we replaced their gender identity with birth sex. This resulted in 378 individuals in the category of man/male (38%) and 616 individuals in the category of woman/female (62.0%). Hereafter, we will refer to this computed variable as gender.

2.3.2. Social network relationship quality variables

The network survey was modified from the Brief Important People Interview (BIPI; Zywiak & Longabaugh, 2002) this approach is commonly used in social network research (Barnett et al., 2019; see Knox et al., 2019 for review). Participants were instructed to identify up to 20 peers who were either important to them, were roommates, or they drank with by selecting their classmates' names from a pull-down list of participants. To be included in the current analyses, participants had to identify at least one peer. For each nominated peer, participants were asked to rate their perceived level of closeness (1 = Not at all close, 5 = Extremely close), frequency of communication (via any means of direct communication; 1 = Not at all frequently, 5 = Extremely frequently), and frequency of drinking together (1 = Not in the past 30 days, 5 = 5 or more times per week). Aggregate or mean scores for each item were computed across all peers for each participant.

2.3.3. Alcohol use

Participants reported their typical number of drinks per drinking occasion in the past 30 days using a single item which was previously used in the 2014 Behavioral Risk Factor Surveillance System Questionnaire (BRFSS; CDC, 2014). Alcohol use among peers was represented by averaging each participant’s nominated peers’ self-reported alcohol quantity.

2.3.4. Participant social media use and ARC sharing

Participants were asked to think of their social media use across all platforms in the past 30 days. Time spent daily was assessed using a single item used previously in Ng Fat et al. (2021), “On average, how much time do you spend on social media?”, with response options from 1 (“Less than once a day) to 8 (“More than 6 hours”). 0 (“I don’t have an account/I don’t use social media”) was also included as an option. If participants endorsed any social media use, they were asked a single question which was modified from Ward et al. (2022) to ask about who they follow rather than how many followers they have, “About how many people or accounts do you follow on the social media site you use the most?” and provided a numerical response. Then they were asked the following question modified from Boyle et al. (2016) to assess self-posting and modified to add publicly or privately to the stem given the importance of this distinction found in Vranken et al. (2022), “Please report the frequency with which you post (i.e., text, pictures, or videos) related to alcohol, drinking, being drunk, or hung-over on social media either publicly or privately (e.g., direct messaging, close friends)” with response options from 0 (“Never”) to 6 (“Daily or almost daily”).

2.3.5. Exposure to ARC

Participants were asked a single question, “What percentage of all of the people or accounts you follow on all your social media display or post alcohol references?” and provided a numerical response. Then they were asked two questions used in prior research (Erevik et al., 2018), “How often do you see posts (i.e., text, images, or videos) on social media either publicly or privately (e.g., direct messaging, close friends) that refer to positive consequences of alcohol use (e.g. increased pleasure, social cohesion, relaxation)?” and separately asked “…that refer to negative consequences of alcohol use (e.g. hangovers, loss of control, hangover anxiety)?” For each nominated peer, participants also were asked a single item “Please report the frequency with which you’ve seen text, pictures, or videos posted by this person related to alcohol, drinking, being drunk, or hung-over on social media either publicly or privately (e.g., direct messaging, close friends) within the past 30 days”. Response options were the same as for participant ARC sharing. For each participant, mean scores across perceived ratings of nominated peers were computed.

2.4. Analytic approach

Social network nominations from the network survey were linked by participant ID (because participants were also nominated peers) such that ties to nominated peers and characteristics of these ties could be analyzed. Network autocorrelation models were used to examine the research questions; these models are essentially regression models with an additional parameter that controls for the nonindependence in the data (Ord, 1975).4 All models controlled for participant gender (SAMHSA, 2022), participant time spent on social media (Ng Fat et al., 2021; Trager et al., 2023), nominated peers’ self-reported alcohol use (Kenney et al., 2017), and participant frequency of ARC sharing (Trager et al., 2023) as these variables have all been found to be related to college drinking in prior research. In Model 1 (RQ1), perceived frequency of exposure to ARC shared by social network members was the predictor and participant drinks per occasion was the outcome. Models 2 through 5 (RQ2) explored moderation of the main effect in Model 1, by including two-way interaction terms between perceived frequency of exposure to ARC shared by network members and perceived closeness, frequency of communication, and frequency of drinking together, respectively. Analyses were conducted in SPSS and in R using the sna package (Butts, 2008).

3. Results

3.1. Drinking and social media behavior descriptive information

Approximately 85.3% of participants consumed one or more alcoholic beverages in the past month. Among drinkers, the average number of drinks per occasion was 2.31 (SD = 1.84). As shown in Table 1, the largest group of participants reported 1-2 hours per day spent on social media. The median number of accounts followed by participants was 500 and participants reported about 13% of the accounts they followed shared ARC. It was common for participants to report perceived exposure to ARC containing positive or negative alcohol consequences, but approximately one-third reported not seeing ARC at all in the past 30 days. Most participants had never shared ARC or had not done so in the past 30 days.

Table 1.

Social Media Descriptive Information

Variable M (SD) or n (%)
Participant Social Media Behaviors M (SD)
 Number of accounts followed 550.32 (438.57)
 Percentage of followed accounts sharing ARC 13.49 (16.37)
Time Spent n (%)
 Don’t have an account/don’t use social media 0 (0.0)
 Less than once a day 36 (3.6)
 1 hour or less a day 160 (16.1)
 1-2 hours a day 305 (30.7)
 2-3 hours a day 259 (26.1)
 3-4 hours a day 136 (13.7)
 4-5 hours a day 65 (6.5)
 5-6 hours a day 22 (2.2)
 More than 6 hours a day 11 (1.1)
Participant Frequency of Sharing ARC
 Never 548 (55.1)
 I’ve shared it before, but not within the past 30 days 283 (28.5)
 Once in the past 30 days 67 (6.7)
 A couple times in the past 30 days 74 (7.4)
 Every week 16 (1.6)
 A couple times a week 5 (0.5)
 Daily or almost daily 1 (0.1)
Frequency of Seeing ARC Refer to Positive Consequences of Alcohol use
 Never 152 (15.3)
 I’ve seen it before, but not within the past 30 days 271 (27.3)
 Once in the past 30 days 123 (12.4)
 A couple times in the past 30 days 260 (26.2)
 Every week 80 (8.1)
 A couple times a week 87 (8.8)
 Daily or almost daily 20 (2.0)
Frequency of Seeing ARC Refer to Negative Consequences of Alcohol use
 Never 252 (25.4)
 I’ve seen it before, but not within the past 30 days 357 (36.0)
 Once in the past 30 days 119 (12.0)
 A couple times in the past 30 days 188 (18.9)
 Every week 40 (4.0)
 A couple times a week 32 (3.2)
 Daily or almost daily 5 (0.5)
Open in a new tab

Note. ARC = alcohol-related content on social media.

Participants on average selected 4.70 network members (SD = 2.93). The average perceived frequency of members sharing ARC was 0.61 (SD = 0.87) which falls between the “0 = never” and “1 = shared it but not in the past 30 days” response options. Relationship closeness was 2.98 (SD = 0.73), frequency of communication was 3.03 (SD = 0.83), and frequency of drinking together was 1.15 (SD = 0.66).

3.2. Frequency of network members sharing ARC and alcohol quantity

As shown in Table 2, perceived frequency of network members sharing ARC, female gender, social network members’ self-reported alcohol use, and participant frequency of sharing ARC were significantly associated with participant typical alcohol quantity. Only participant time spent on social media was not associated with alcohol quantity5.

Table 2.

Associations Between Participant and Network Member Characteristics, ARC Sharing and Participant Alcohol Quantity

Variables Alcohol Quantity
B SE p
Model 1: SN Frequency of ARC Sharing
 Participant Gender −0.38 ** 0.09 <.001
 Participant Time Spent on Social Media 0.03 0.03 .41
 Participant Frequency of ARC Sharing 0.17 ** 0.05 <.001
 Number of SN Members Selected 0.04 * 0.01 .01
 SN Members’ Self-Report of Alcohol Quantity 0.79 ** 0.03 <.001
 Participant Perception of SN Member Frequency of ARC Sharing 0.20 * 0.06 .002
Model 2: SN Frequency of ARC Sharing x Participant Gender
 Participant Gender −0.39 ** 0.09 <.001
 Participant Time Spent on Social Media 0.03 0.03 .40
 Participant Frequency of ARC Sharing 0.17 ** 0.05 <.001
 Number of SN Members Selected 0.04 * 0.01 .01
 SN Members’ Self-Report of Alcohol Quantity 0.79 ** 0.03 <.001
 Participant Perception of SN Member Frequency of ARC Sharing 0.31 * 0.09 <.001
 SN Frequency of ARC Sharing x Participant Gender −0.18 0.10 .08
Model 3: SN Frequency of ARC Sharing x SN Average Closeness
 Participant Gender −0.59 ** 0.10 <.001
 Participant Time Spent on Social Media 0.05 0.04 .19
 Participant Frequency of ARC Sharing 0.08 0.05 .10
 Number of SN Members Selected 0.01 0.02 .64
 SN Members’ Self-Report of Alcohol Quantity 0.62 ** 0.04 <.001
 Participant Perception of SN Member Frequency of ARC Sharing 0.22 * 0.06 <.001
 SN Average Closeness 0.13 0.07 .07
 SN Frequency of ARC Sharing x SN Average Closeness 0.04 0.08 .67
Model 4: SN Frequency of ARC Sharing x SN Average Frequency of Communication
 Participant Gender −0.62 ** 0.10 <.001
 Participant Time Spent on Social Media 0.05 0.04 .19
 Participant Frequency of ARC Sharing 0.08 0.05 .08
 Number of SN Members Selected 0.01 0.02 .55
 SN Members’ Self-Report of Alcohol Quantity 0.61 ** 0.04 <.001
 Participant Perception of SN Member Frequency of ARC Sharing 0.19 * 0.06 .003
 SN Average Frequency of Communication 0.19 * 0.06 .003
 SN Frequency of ARC Sharing x SN Average Frequency of Communication 0.12 0.07 .09
Model 5: SN Frequency of ARC Sharing x Average Frequency of Drinking with SN
 Participant Gender −0.60 ** 0.10 <.001
 Participant Time Spent on Social Media 0.04 0.04 .22
 Participant Frequency of ARC Sharing 0.07 0.05 .12
 Number of SN Members Selected −0.01 0.02 .49
 SN Members’ Self-Report of Alcohol Quantity 0.52 ** 0.04 <.001
 Participant Perception of SN Member Frequency of ARC Sharing 0.17 * 0.07 .01
 Average Frequency of Drinking with SN 0.41 ** 0.08 <.001
 SN Frequency of ARC Sharing x Average Frequency of Drinking with SN 0.06 0.08 .43
Open in a new tab

Note. ARC = alcohol-related content on social media, SN = social network. Participant gender had the following response options: 0 = male, 1 = female. Social media checking frequency had the following response options: 0 (“I don’t have an account/I don’t use social media”), 1 (“Less than once a day) to 2 (“1 hour or less a day”), 3 (“1-2 hours a day”), 4 (“2-3 hours a day”), 5 (“3-4 hours a day”), 6 (“4-5 hours a day”), 7 (“5-6 hours a day”), and 8 (“More than 6 hours”). Bold values are significant.

*

p < .05

**

p < .001.

3.3. Moderation analyses

Across all models, female gender was negatively associated with participant alcohol quantity while social network members’ self-report of alcohol quantity, and perceptions of social network members’ frequency of ARC sharing were positively associated with participant alcohol quantity. The number of social network members selected, and participant ARC sharing were significantly associated with alcohol quantity in some but not all models (detailed below). Participant time spent on social media was not associated with alcohol quantity.

3.3.1. Gender.

The number of social network members selected and participant frequency of ARC sharing were positively associated with participant alcohol quantity. However, gender did not significantly interact with frequency of social network ARC sharing to predict alcohol quantity suggesting a lack of moderation.

3.3.2. Closeness.

The number of social network members selected and participant sharing of alcohol content were not associated with their alcohol quantity. Average closeness with the social network was not associated with alcohol quantity and did not interact with social network ARC sharing to predict quantity suggesting a lack of moderation.

3.3.3. Frequency of communication.

The number of social network members selected and participant sharing of alcohol content were not associated with their alcohol quantity. Frequency of communication was associated with participant alcohol quantity but did not interact with social network ARC sharing to predict participant alcohol quantity.

3.3.4. Drinking frequency with network members.

The number of social network members selected and participant sharing of alcohol content were not associated with quantity. Frequency of drinking with social network members was associated with participant alcohol quantity but did not interact with social network ARC sharing frequency to predict participant alcohol quantity.

4. Discussion

The aims of this study were to determine whether greater perceived frequency of exposure to ARC shared by social network members (i.e., nominated peers) was associated with participant alcohol use and whether this hypothesized effect was moderated by participant gender and specific perceived relationship qualities with social network members. We found that greater perceived frequency of exposure to ARC was associated with participant alcohol consumption over and above participant frequency of sharing ARC and network members’ self-report of their own drinking. Higher perceived frequency of communication with network members, and higher perceived frequency of drinking with network members were positively related to participant drinking while female sex was negatively associated, but none of the investigated characteristics moderated the effect of perceived ARC sharing by nominated peers on participant drinking.

Our finding that greater perceived frequency of exposure to network member ARC was associated with greater participant drinking is consistent with previous research (Boyle et al., 2016; Davis et al., 2021). However, most studies examining ARC exposure use a global approach asking about a reference group as a whole rather than specific individuals (for a review see Strowger & Braitman, 2022). A strength of the current study was that we asked participants to name important individuals in their lives and describe their behaviors and qualities of their relationships. Although we aggregated responses across each participant’s network, participants may find it easier to accurately recall information about specific people than their friend group as a whole. The current study findings reinforce that perceived frequency of ARC exposure is an important factor in alcohol use when controlling for key variables such as participant sharing of ARC and peers’ self-reported drinking; indicating that peer ARC exposure may have an influence on drinking that is independent of other variables.

We did not find that participant gender moderated the association between perceived frequency of exposure to ARC and alcohol use which is inconsistent with prior research that found that sex moderates associations between ARC exposure and alcohol use among first-year students (Boyle et al., 2016; Davis et al., 2021; Roberson et al., 2018). Importantly, these prior studies assessed ARC exposure from peers broadly defined whereas our study assessed ARC exposure from important peers, roommates, and those they drank with. Given that proximal influences are more impactful than distal ones on drinking behavior (Borsari & Carey, 2003), it is possible that what matters most is the ARC source, close peers, rather than the gender of the viewer.

The lack of moderation of the three relationship qualities could be related to the fact that the network was narrow; participants were only allowed to select individuals in their same class year. This instruction may have limited the average and variability in ratings of relationship characteristics among peers (the average score was 3 on a scale 1-5), and likely did not capture all important and influential friends. The moderation outcomes might have differed if participants’ networks had included friends outside the class-year network. Further, although these relationship qualities have been found to moderate the association between peer drinking and college alcohol use, they may not be as relevant to exposure to peer ARC. It could be instead that what is portrayed in ARC shared by peers including featuring the participant who is viewing the content or an event they attended may be more personally relevant and influential. Future studies should include both quantitative and qualitative work to understand what aspects of peer ARC are most influential for personal drinking behavior.

4.1. Limitations & future directions

The study findings must be interpreted in the context of study limitations. First, the cross-sectional nature of this work limits causal interpretations. Research examining changes over time in exposure to ARC and personal alcohol use would be valuable in furthering our understanding of these relationships, particularly studies using ecological momentary assessment or daily diary designs that could reveal the proximal impact of such exposures. Second, there are advantages and disadvantages to different methods for deriving computed social network variables. We computed an aggregate value of alcohol use across nominated peers derived from their own self-report. However, for ARC sharing, we relied on participant perceptions of their nominated peers. A recent study by LaBrie et al. (2021) examined discrepancies between self-reported and objectively-measured sharing and viewing ARC. Results indicated that perceptions of social media use differ significantly from observed behaviors and that perceptions may be more influential than observed rates of sharing or viewing (Geusens & Beullens, 2021; LaBrie et al., 2021).

Future work should examine if there are discrepancies between participant perceptions of sharing by their network members, network member self-reports, and observed ARC frequency and how these discrepancies relate to personal drinking. We also aggregated the relationship characteristics across network members; a more granular analysis at the dyadic level examining how these characteristics predict discrepancies in network member and participant reports of ARC sharing may reveal different findings. We did not examine how the effects ARC exposure on specific social media platforms affect drinking because social media use is increasingly becoming diverse. Finally, we did not use a global question to collect information on exposure to ARC shared by close friends, precluding our ability to examine whether exposure to ARC from specific peers is more strongly associated with drinking behavior than global exposure to friend ARC. Prior research found that perceptions of individual peer drinking levels are more strongly associated with participant drinking than global perceptions of close friend drinking (Russell et al., 2020) and lend support to this assumption, but future research should directly compare global and specific ARC exposure.

4.2. Implications & conclusions

This study makes a significant contribution to the literature as it is the first to explore the association between exposure to ARC—posted on social media by close peers—and young adults’ alcohol use using a sociocentric network approach. The finding that greater perceived frequency of ARC sharing by important peers was associated with alcohol quantity has implications for improving the efficacy of existing college drinking interventions. College drinking interventions that include personalized feedback about risks related to alcohol use may wish to include content on alcohol-related social media posts. For example, participants perceived seeing ARC depicting positive consequences more often than negative consequences. Content that mostly shows positives of drinking both glamorizes and normalizes drinking behavior. College students who have little to no experience with alcohol may view this content and not be adequately cognizant of alcohol-related negative consequences, while for heavier drinkers exposure to this content may reinforce their alcohol cognitions and behaviors. Including ARC posting norms and teaching skills to critically evaluate peer ARC (Dunn et al., 2020) in existing college drinking interventions may help to reduce the influence of ARC on behavior. Similarly, clinicians working one-on-one with young adults to reduce their alcohol use may wish to discuss exposure to ARC on social media shared by peers as another relevant source of peer influence on their behavior. Diving deeper into understanding the impact of exposure to this content on their alcohol use will enable clinicians to initiate a dialogue on the most effective ways to mitigate the effects through critical thinking and media literacy.

Highlights.

  • Perceived exposure to ARC from specific peers was assessed

  • Greater exposure to network member ARC was associated with higher alcohol quantity in participants

  • Participant sex and relationship qualities were not significant moderators

  • Peer ARC exposure independently predicts college drinking above and beyond peer drinking

  • Including ARC media literacy in alcohol interventions may reduce the effects of exposure

Acknowledgments

Megan Strowger is currently supported by an institutional training grant (T32AA007459; PI: Miranda) and was previously supported by an individual fellowship award from the National Institute of Alcohol Abuse and Alcoholism (F31AA029945; PI: Strowger). In addition, this research was supported in part by grant numbers K01AA025994 (PI: Meisel), K08AA029181 (PI: Haikalis), and R01AA023522 (PI: Barnett). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Alcohol Abuse and Alcoholism.

Role of funding sources

Funding for this study was provided by NIAAA grant numbers T32AA007459, F31AA029945, K01AA025994, K08AA029181, and R01AA023522. NIAAA had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CRediT authorship contribution statement

Megan Strowger: Conceptualization, Writing – original draft, Writing – review & editing. Matthew K. Meisel: Conceptualization, Formal analysis, Data curation, Writing – original draft, Writing – review & editing. Michelle Haikalis: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Michelle Rogers: Conceptualization, Data curation, Writing – review & editing. Nancy P. Barnett: Conceptualization, Funding Acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

1

Students were eligible for the research if they were a currently enrolled student from the original class of 2021; this included those who might have taken a leave so were not in their third year and students who were on track to graduate early.

2

Students who were initially eligible at baseline were allowed to join the study at any assessment; this is common for sociocentric network surveys as it allows for the most complete representation of the network and ties between participants.

3

Of those who completed 1,173 selected at least one social network member (92.7%).

4

In standard regression, observations are assumed to be independent. However, network autocorrelation models account for the inherent dependence and correlation among observations (i.e., participants can be nominated peers and nominated peers can be participants). Failure to account for this can result in underestimating standard errors and/or committing a Type I error. In network autocorrelation models, we employ a model: y = λWy + Xβ + ε. Here, y represents the dependent variable, λ is the autocorrelation parameter (zero when there is no dependence), W is a network-based matrix, X is a vector of covariates, β is a vector of coefficients, and ε signifies independent errors. Additionally, y is on both sides of the model because participants can be nominated peers and nominated peers can be participants. When λ is zero (i.e., there is no dependence between units in the network), the model simplifies to standard linear regression. Network autocorrelation models use maximum likelihood estimation.

5

In a set of sensitivity analyses, the main models were rerun controlling for the proportion of social network members who were friends versus other relationship types and the proportion of females in the network versus males. The patterns of findings did not change, and these covariates were non-significant across all models.

There are no known conflicts of interest to disclose.

References

  1. Bandura A, & Walters RH (1977). Social learning theory (Vol. 1). Prentice Hall: Englewood cliffs. [Google Scholar]
  2. Barnett NP, Clark MA, Kenney SR, DiGuiseppi G, Meisel MK, Balestrieri S, Ott MQ, & Light J (2019). Enrollment and assessment of a first-year college class social network for a controlled trial of the indirect effect of a brief motivational intervention. Contemporary Clinical Trials, 76, 16–23. 10.1016/j.cct.2018.10.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Borsari B, & Carey KB (2001). Peer influences on college drinking: A review of the research. Journal of Substance Abuse, 13(4), 391–424. 10.1016/S0899-3289(01)00098-0 [DOI] [PubMed] [Google Scholar]
  4. Borsari B, & Carey KB (2003). Descriptive and injunctive norms in college drinking: A meta-analytic integration. Journal of Studies on Alcohol, 64(3), 331–341. 10.15288/jsa.2003.64.331 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Borsari B, & Carey KB (2006). How the quality of peer relationships influences college alcohol use. Drug and Alcohol Review, 25(4), 361–370. 10.1080/09595230600741339 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boyle SC, LaBrie JW, Froidevaux NM, & Witkovic YD (2016). Different digital paths to the keg? How exposure to peers' alcohol-related social media content influences drinking among male and female first-year college students. Addictive Behaviors, 57, 21–29. 10.1016/j.addbeh.2016.01.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Butts CT (2008). “Social Network Analysis with sna.” Journal of Statistical Software, 24(6). 1–51. 10.18637/jss.v024.i06 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Carolan BV (2014). Social network analysis and education: Theory, methods & applications. SAGE Publications, Inc., 10.4135/9781452270104 [DOI] [Google Scholar]
  9. Centers for Disease Control and Prevention (CDC) Behavioral Risk Factor SurveillanceSystem Survey Questionnaire, US Department of Health Human Services, Centers for Disease Control and Prevention, Atlanta, GA: (2014) https://www.cdc.gov/brfss/questionnaires/pdf-ques/2014_BRFSS.pdf [Google Scholar]
  10. Cirillo MN, Halbert JP, Smith JG, Alamiri NS, & Ingersoll KS (2022). # BingeDrinking—Using Social Media to Understand College Binge Drinking: Qualitative Study. JMIR Human Factors, 9(2), e36239. 10.2196/36239 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Corcoran E, Doucette H, Merrill JE, Pielech M, López G, Egbert A, Nelapati S, Gabrielli J, Colby SM, & Jackson KM (2023). A qualitative analysis of adolescents' perspectives on peer and influencer alcohol-related posts on social media. Drug and Alcohol Review, 10.1111/dar.13720 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Conway JM, & DiPlacido J (2015). The indirect effect of alcohol use on GPA in first-semester college students: The mediating role of academic effort. Journal of College Student Retention: Research, Theory & Practice, 17(3), 303–318. 10.1177/1521025115575705 [DOI] [Google Scholar]
  13. Cruz JE, Emery RE, & Turkheimer E (2012). Peer network drinking predicts increased alcohol use from adolescence to early adulthood after controlling for genetic and shared environmental selection. Developmental Psychology, 48(5), 1390–1402. 10.1037/a0027515 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Curtis BL, Lookatch SJ, Ramo DE, McKay JR, Feinn RS, & Kranzler HR (2018). Meta-analysis of the association of alcohol-related social media use with alcohol consumption and alcohol-related problems in adolescents and young adults. Alcohol: Clinical and Experimental Research, 42(6), 978–986. 10.1111/acer.13642 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Davis JP, Christie NC, Lee D, Saba S, Ring C, Boyle S, Pedersen ER, & LaBrie J (2021). Temporal, sex-specific, social media-based alcohol influences during the transition to college. Substance Use & Misuse, 56(8), 1208–1215. 10.1080/10826084.2021.1914106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dunn HK, Pearlman DN, & Florin P (2020). A pilot study exploring the viability of focus groups as a brief social media literacy intervention for substance use prevention among adolescents. Journal of Adolescent and Family Health, 11(1), 6. https://scholar.utc.edu/jafh/vol11/iss1/6 [Google Scholar]
  17. Erevik EK, Pallesen S, Andreassen CS, Vedaa Ø, & Torsheim T (2018). Who is watching user-generated alcohol posts on social media?. Addictive Behaviors, 78, 131–137. 10.1016/j.addbeh.2017.11.023 [DOI] [PubMed] [Google Scholar]
  18. Finlay AK, Ram N, Maggs JL, & Caldwell LL (2012). Leisure activities, the social weekend, and alcohol use: Evidence from a daily study of first-year college students. Journal of Studies on Alcohol and Drugs, 73(2), 250–259. 10.15288/jsad.2012.73.250 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Geusens F, & Beullens K (2021). Perceptions surpass reality: Self-reported alcohol-related communication on Instagram is more strongly related with frequency of alcohol consumption and binge drinking than actual alcohol-related communication. Drug and Alcohol Dependence, 227, 109004. 10.1016/j.drugalcdep.2021.109004 [DOI] [PubMed] [Google Scholar]
  20. Gupta H, Pettigrew S, Lam T, & Tait RJ (2016). A systematic review of the impact of exposure to internet-based alcohol-related content on young people's alcohol use behaviours. Alcohol and Alcoholism, 51(6), 763–771. 10.1093/alcalc/agw050 [DOI] [PubMed] [Google Scholar]
  21. Hendriks H, Van den Putte B, Gebhardt WA, & Moreno MA (2018). Social drinking on social media: content analysis of the social aspects of alcohol-related posts on Facebook and Instagram. Journal of Medical Internet Research, 20(6), e226. 10.2196/jmir.9355 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Huang GC, Soto D, Fujimoto K, & Valente TW (2014). The interplay of friendship networks and social networking sites: Longitudinal analysis of selection and influence effects on adolescent smoking and alcohol use. American Journal of Public Health, 104(8), e51–e59. 10.2105/AJPH.2014.302038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Huang GC, Unger JB, Soto D, Fujimoto K, Pentz MA, Jordan-Marsh M, & Valente TW (2014). Peer influences: the impact of online and offline friendship networks on adolescent smoking and alcohol use. Journal of Adolescent Health, 54(5), 508–514. 10.1016/j.jadohealth.2013.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kenney SR, Ott M, Meisel MK, & Barnett NP (2017). Alcohol perceptions and behavior in a residential peer social network. Addictive Behaviors, 64, 143–147. 10.1016/j.addbeh.2016.08.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kiciman E, Counts S, & Gasser M (2018, June 25-28). Using longitudinal social media analysis to understand the effects of early college alcohol use [Paper presentation]. ICWSM 2018: Palo Alto, CA, United States. https://www.aaai.org/ocs/index.php/ICWSM/ICWSM18/paper/view/17844/17008 [Google Scholar]
  26. Knox J, Schneider J, Greene E, Nicholson J, Hasin D, & Sandfort T (2019). Using social network analysis to examine alcohol use among adults: A systematic review. PloS One, 14(8), e0221360. 10.1371/journal.pone.0221360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. LaBrie JW, Trager BM, Boyle SC, Davis JP, Earle AM, & Morgan RM (2021). An examination of the prospective associations between objectively assessed exposure to alcohol-related Instagram content, alcohol-specific cognitions, and first-year college drinking. Addictive Behaviors, 119, 106948. 10.1016/j.addbeh.2021.106948 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lac A, & Donaldson CD (2018). Testing competing models of injunctive and descriptive norms for proximal and distal reference groups on alcohol attitudes and behavior. Addictive Behaviors, 78, 153–159. 10.1016/j.addbeh.2017.11.024 [DOI] [PubMed] [Google Scholar]
  29. Lau-Barraco C, Braitman AL, Leonard KE, & Padilla M (2012). Drinking buddies and their prospective influence on alcohol outcomes: Alcohol expectancies as a mediator. Psychology of Addictive Behaviors, 26(4), 747–758. 10.1037/a0028909 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mason MJ, Zaharakis NM, Rusby JC, Westling E, Light JM, Mennis J, & Flay BR (2017). A longitudinal study predicting adolescent tobacco, alcohol, and cannabis use by behavioral characteristics of close friends. Psychology of Addictive Behaviors, 31(6), 712–720. 10.1037/adb0000299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Moreno MA, D'Angelo J, Hendriks H, Zhao Q, Kerr B, & Eickhoff J (2021). A prospective longitudinal cohort study of college students' alcohol and abstinence displays on social media. Journal of Adolescent Health, 69(3), 440–446. 10.1016/j.jadohealth.2021.01.027 [DOI] [PubMed] [Google Scholar]
  32. Ng Fat L, Cable N, & Kelly Y (2021). Associations between social media usage and alcohol use among youths and young adults: findings from Understanding Society. Addiction, 116(11), 2995–3005. 10.1111/add.15482 [DOI] [PubMed] [Google Scholar]
  33. Ord K (1975). Estimation methods for models of spatial interaction. Journal of the American Statistical Association, 70(349), 120–126. 10.1080/01621459.1975.10480272 [DOI] [Google Scholar]
  34. Patrick ME, Terry-McElrath YM, Evans-Polce RJ, & Schulenberg JE (2020). Negative alcohol-related consequences experienced by young adults in the past 12 months: Differences by college attendance, living situation, binge drinking, and sex. Addictive Behaviors, 105, 106320. 10.1016/j.addbeh.2020.106320 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Reifman A, Watson WK, & McCourt A (2006). Social networks and college drinking: Probing processes of social influence and selection. Personality and Social Psychology Bulletin, 32, 820–832. 10.1177/0146167206286219 [DOI] [PubMed] [Google Scholar]
  36. Rinker DV, Krieger H, & Neighbors C (2016). Social network factors and addictive behaviors among college students. Current Addiction Reports, 3(4), 356–367. 10.1007/s40429-016-0126-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Roberson AA, McKinney C, Walker C, & Coleman A (2018). Peer, social media, and alcohol marketing influences on college student drinking. Journal of American College Health, 66(5), 369–379. 10.1080/07448481.2018.1431903 [DOI] [PubMed] [Google Scholar]
  38. Russell AM, Barry AE, & Patterson MS (2020). A comparison of global and egocentric network approaches for assessing peer alcohol use among college students in the United States. Drug and Alcohol Review, 39(7), 984–993. 10.1111/dar.13140 [DOI] [PubMed] [Google Scholar]
  39. SAMHSA, Center for Behavioral Health Statistics and Quality. 2022 National Survey on Drug Use and Health. Table 2.27B—Alcohol use in past month: among people aged 12 or older; by age group and demographic characteristics, percentages, 2021 and 2022. [cited 2023 Dec 8]. Available from: https://www.samhsa.gov/data/sites/default/files/reports/rpt42728/NSDUHDetailedTabs2022/NSDUHDetailedTabs2022/NSDUHDetTabsSect2pe2022.htm#tab2.27b [Google Scholar]
  40. Schulenberg JE, Patrick ME, Johnston LD, O’Malley PM, Bachman JG, & Miech RA (2021). Monitoring the Future national survey results on drug use, 1975–2020: Volume II, College students and adults ages 19–60. Ann Arbor: Institute for Social Research, The University of Michigan. Available at http://monitoringthefuture.org/pubs.html#monographs [Google Scholar]
  41. Steers MLN, Neighbors C, Wickham RE, Petit WE, Kerr B, & Moreno MA (2019). My friends, I’m# SOTALLYTOBER: A longitudinal examination of college students’ drinking, friends’ approval of drinking, and Facebook alcohol-related posts. Digital Health, 5, 1–11. DOI: 10.1177/2055207619845449 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Strowger M, Ayala Guzman R, Geyer RB, Ward RM, & Braitman AL (2023). Following social media influencers who share alcohol-related content is associated with college drinking. Drug and Alcohol Review. 10.1111/dar.13694 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Strowger M, & Braitman AL (2023). Using social network methodology to examine the effects of exposure to alcohol-related social media content on alcohol use: A critical review. Experimental and Clinical Psychopharmacology, 31(1), 280–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Strowger M, Braitman AL, & Barnett NP (2022). The association between social network members sharing alcohol-related social media content and alcohol outcomes among college student drinkers. Alcohol: Clinical and Experimental Research, 46(8), 1592–1602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Trager BM, Morgan RM, Boyle SC, Ishino FAM, & LaBrie JW (2023). Capturing the bigger picture: A gestalt of general and alcohol-specific social media usage during the transition to college as a predictor of first-year alcohol use and consequences. Addictive Behaviors, 136, 107472. 10.1016/j.addbeh.2022.107472 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Vranken S, Kurten S, & Beullens K (2022). The temporality and accessibility of message types (TAMT) model: Examining social media message types and the associations between exposure to alcohol and binge drinking. Cyberpsychology-Journal Of Psychosocial Research On Cyberspace, 16(5), Article 4. 10.5817/CP2022-5-4. [DOI] [Google Scholar]
  47. Ward RM, Dumas TM, Lewis MA, & Litt DM (2022). Likelihood of Posting Alcohol-Related Content on Social Networking Sites–Measurement Development and Initial Validation. Substance Use & Misuse, 57(7), 1111–1119. 10.1080/10826084.2022.2064505 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zywiak W, Longabaugh R The Brief Important People Interview Manual, Center for Alcohol and Addiction Studies, Providence, RI: (2002) [Google Scholar]

RESOURCES