About this paper symposium

---

Panel information
Panel 22. Social Relationships

Paper #1
School Structure Explains Age Group Differences in Friend Selection Similarity
Author information	Role
Brett Laursen, Ph.D., Florida Atlantic University, United States	Presenting author
Molly Selover, Florida Atlantic University, United States	Non-presenting author
Mary Page Leggett-James, Florida Atlantic University, United States	Non-presenting author
Goda Kaniušonytė, Mykolas Romeris University, Lithuania	Non-presenting author
Abstract
The origins of age-related increases in friend similarity can be traced to two (not mutually exclusive) possibilities (Laursen, 2017): Youth increasingly select similar others as friends and youth increasingly influence one another to become more similar. The present study explores whether selection drives increases in friend similarity in the USA because the pool of similar others who could potentially serve as friends expands between primary and middle school. In contrast, in cultures such as Lithuania, selection similarity should not increase because the number of classmates remains largely unchanged from primary to middle school. Participants included 204 (93 boys, 111 girls) students attending 2 public schools in Florida (USA), and 339 (182 boys, 157 girls) students attending 7 public schools in Lithuania. Florida school records indicated that 42.9% were European-American, 33.9% were Hispanic-American, 14.3% were African-American, and 8.9% were mixed and other backgrounds. Lithuania students were nearly all ethnic Lithuanian. Florida primary school students (grades 4-5) attended classes with the same classmates (M=22.7); middle school students (grades 6-7) attended classes with a rotating group of students (M=82.7) drawn from 4 homerooms, referred to as pods. Lithuania primary school (grade 4) and middle school (grades 5-7) students attended classes with the same classmates (M=15.4 and 17.4, respectively). Students completed peer nomination surveys twice during a single academic year, from which future friend dyads [i.e., partners who reciprocally nominated one another as friends in January (Time 2) but not in November (Time 1)] were identified (Florida N=168; Lithuania N=229). Students also completed a peer nomination inventory to identify classmates/podmates who best fit the following descriptors: academic achievement, accepted, athletic, disruptive, popular, and rejected. Within-dyad intraclass (interpreted as r2) correlation contrasts (Schenker & Gentleman, 2001) tested the hypothesis that future friend similarity in Florida is greater for middle school students than primary school students because the former have more options (i.e., interact with a larger number of classmates over the course of the day) to identify compatible companions. No differences were anticipated between Lithuania primary school and middle school students, because each have a similar number of classmates. We present partial correlations, controlling for the number of each participant’s future friends. Confidence interval contrasts indicated that Florida middle school future friends were more similar than Florida primary school future friends on academic achievement, acceptance, athleticism, and popularity (see Figure 1). In contrast, Lithuania middle school future friends were not more similar than primary school future friends on any variable; primary school friends were more similar than middle school friends on disruptiveness (see Figure 2). The same pattern of results emerged when Lithuania 4th/5th graders were contrasted with 6th/7th graders. Strong arguments have been made that increases in friend similarity are a product of increases in susceptibility to peer influence (Laursen & Faur, 2022). Often overlooked is a more benign possibility: That heightened similarity during middle school can also be traced to changes in school structure that increase exposure to a greater variety of classmates, enhancing opportunities to identify similar others as friends.

Paper #2
The relative effects of two peer influence mechanisms on cigarette use onset and frequency
Author information	Role
Nathan Field, Ph.D., University of Virginia, United States	Presenting author
Meghan Costello, Harvard Medical School, United States	Non-presenting author
Lauren Breeden, University of Virginia, United States	Non-presenting author
Joseph P. Allen, University of Virginia, United States	Non-presenting author
Abstract
Decades of psychological and epidemiological research suggests that adolescents’ peers’ smoking attitudes (Field & Prinstein, 2023) and behaviors (Alexander et al., 2001) are among the most robust determinants of adolescents’ own smoking initiation and frequency. However, it is unclear which peers may be most influential towards adolescents’ cigarette use, and whether the routes of transmission may differ for best friends’ influence compared to the larger peer group’s influence. Further, the majority of work in this domain has relied on 1) perceived smoking attitudes and behaviors and 2) the influence to begin or continue smoking rather than abstaining or reducing smoking frequency. To provide some clarity to these questions, the current study examined (1) how close friends’ actual smoking behavior and (2) their peer groups’ influence to abstain from smoking are simultaneously associated with adolescents’ own smoking initiation and frequency over a 5-year period throughout ages 13 – 17. Participants were from a multi-reporter longitudinal study of social development. From ages 13-17, participants (N=184; 99 female; 58% White, 29% Black, 13% other racial identities) completed self-report surveys and nominated their best friend to be included in survey measures with them throughout the study. Participants reported annually on their own cigarette use frequency, as well as their peer group’s influence not to smoke, while participants’ best friends reported on their own cigarette use. Multivariate discrete time survival analyses examined how best friends’ cigarette use initiation and perceived peer influence towards not smoking were prospectively associated with participants’ own odds of smoking initiation. A bivariate latent curve model with a time-varying covariate (see Figure 1) examined how best friends’ smoking frequency and peer influence towards not smoking were simultaneously associated with the rate of change in smoking frequency across development. Survival analyses revealed that best friend’s cigarette use was significantly associated with a heightened probability of adolescents’ own smoking onset. Compared to youth whose friends did not ever smoke cigarettes, youth with friends who smoke are roughly three times more likely to begin smoking themselves (see Table 1). In contrast, friends’ influence to not smoke did not significantly reduce the likelihood of adolescent smoking onset (see Table 1). The latent curve model revealed a significant negative relation between the slope for cigarette smoking and the slope for friends’ influence not to smoke (b = -0.31, p = 0.036) suggesting that the more slowly friends lose influence over time, the slower the rate of increase in cigarette smoking becomes. There were additional concurrent (b = 0.19, p < .001) and one-year-lagged effects (b = 0.27, p < .001) between adolescents’ best friends’ cigarette consumption and their own smoking behavior. These effects reveal that best friends’ smoking behavior is both concurrently and longitudinally associated with adolescents’ own smoking frequency. Results suggest that neither knowing the specific best friendship norms and behaviors nor the greater peer group norms is sufficient, but rather both additively contribute to shape adolescents’ cigarette use patterns. Each also has unique routes of transmission that should be the focus of future empirical work.

Paper #3
Characterizing the Social Landscape of Nicotine Vaping Among Adolescents Attempting to Quit Vaped Nicotine
Author information	Role
Meghan Costello, Ph.D., Harvard Medical School, United States	Presenting author
Bryn Evohr, Massachusetts General Hospital, United States	Non-presenting author
Jay Dufour, Massachusetts General Hospital, United States	Non-presenting author
Vanessa Iroegbulem, Massachusetts General Hospital, United States	Non-presenting author
Kelly Casottana, Massachusetts General Hospital, United States	Non-presenting author
Gladys Pachas, Massachusetts General Hospital, United States	Non-presenting author
Cori Cather, Massachusetts General Hospital/Harvard Medical Schoo, United States	Non-presenting author
Randi Schuster, Massachusetts General Hospital/Harvard Medical School, United States	Non-presenting author
A. Eden Evins, Massachusetts General Hospital/Harvard Medical School, United States	Non-presenting author
Abstract
Current (past 3-day) vaped nicotine use among adolescents and young adults has risen from 1.5% to 20.8% in the past decade, with lifetime use estimates exceeding 40% in the United States in 2019 (Hammond et al., 2020). Vaped nicotine use is associated with increased risk for addiction to nicotine and other substances (Curran et al., 2018), mental health symptoms (Tervo-Clemmens et al., 2024), and chronic physical illness (Winnickoff et al., 2024). Peer use of nicotine has been indicated as a major predictor of nicotine use in adolescents, particularly in early stages of nicotine use (Sweeney & Sharmin, 2020). We hypothesize that peer vape use is a barrier to quitting vaping among adolescents with nicotine use disorder participating in a randomized control trial (RCT) to support quitting vaping. Participants include 300 adolescents and young adults (age 16-25) enrolled in a 12-week, parallel-group, placebo-controlled RCT of a cognitive behavior therapy (CBT) intervention and pharmacotherapy (varenicline) for quitting vaped nicotine. All participants endorsed nicotine dependence at study start, as indicated by daily/near daily vaping for at least the past three months and a score of ≥4 on the E-cigarette Dependence Inventory (ECDI; Pienkowski et al., 2022). Participants reported perceived number of peers who use vaped nicotine at baseline. They reported frequency of vape use (estimated days per week used, 0-7) for each of their 10 closest friends at baseline, end-of-intervention, and 3-month follow-up. Past 7-day nicotine abstinence was biochemically-verified via urinalysis. At baseline, participants reported that many of their peers used vaped nicotine (M=6.9 peers, SD=6.8, Range=0-60 peers), and 50% reported their best friend vaped nicotine at least one day per week (M=4 days/week, SD=3.3 days/week; see Figure 1). A series of multi-level mixed-effects models evaluated associations from peer vape use to participant nicotine abstinence at end-of-intervention and follow-up, accounting for treatment condition (see Table 1). Number of peers using vaped nicotine at baseline was associated with nicotine abstinence at intervention conclusion. Adolescents who reported higher baseline number of friends using vaped nicotine were less likely to achieve nicotine abstinence at end-of-intervention (beta=-.11, p=.03). Regarding reported vaping frequency among adolescents’ best friends: Best friend’s end-of-intervention vape frequency was negatively associated with participant abstinence at end-of-intervention (beta=-.25, p=.05). Best friend’s vape use at both baseline (beta=-.35, p=.01) and 3-month follow-up (beta=-.32, p=.003) was negatively associated with abstinence at 3-month follow-up. Regarding mean reported vape use among adolescents’ close peer network: Abstinence at end-of-intervention was negatively associated with mean vaped nicotine frequency at baseline (beta=-.33, p=.03) and end-of-intervention (beta=-.23, p=.05). Peer network vape frequency at end-of-intervention was also negatively associated with abstinence status at 3-month follow-up (beta=-.32, p=.01). Even among motivated young people (i.e., enrolled in a quit-vaping intervention study), the presence of close others who regularly use vaped nicotine poses a barrier to quitting. To leverage the power of peer influence, perhaps partnered interventions may support the maintenance of abstinence, as they have in other mental health treatments (Baucom et al., 2015).

⇦ Back to session