How Friends Influence Teens' Long-Term Drug and Alcohol Use

According to the Centers for Disease Control and Prevention, drug overdose deaths in the United States were four times more prevalent in 2018 than in 1999, and drug overdose deaths increased by 30 percent during the COVID-19 pandemic. Many factors, including stress, mental health problems, and individuals’ perceptions of norms in their social circles, to name a few, contribute to substance use and misuse. Substance use during adolescence generally begins in social situations with peers.

Previous research has provided evidence for both selection and socialization in relations between individuals’ own substance use and their peers’ substance use. That is, individuals are more likely to choose to be friends with people who are similar to themselves in their use of alcohol, marijuana, and other substances. In addition, friends influence each other’s substance use over time so they become more similar to each other in use.

My colleagues and I studied how individuals’ perceptions of their peers’ substance use and reports of their own substance use predict one another across development from adolescence to adulthood. We recruited 1,048 children from two samples. One was a low-risk community sample, and the other was a sample at higher risk for behavior problems, as rated by elementary school teachers and parents. The children originally attended schools in Indiana, North Carolina, Pennsylvania, Tennessee, and Washington, and we followed the samples through periodic interviews from childhood to adulthood. In early adolescence (age 12 or 13), middle adolescence (age 16 or 17), and early adulthood (age 20 or 22), participants each reported on their own and their peers’ substance use. Participants also reported on their own substance use at age 32 or 34.

We found stability in individuals’ substance use from early to middle adolescence and from adolescence to early adulthood, meaning that adolescents who used more substances relative to their peers in early adolescence continued to do so later in adolescence and in early adulthood, even if their overall level of substance use increased. At each point in development, individuals’ perceptions of their peers’ substance use were correlated with individuals’ own substance use. These findings suggest that one way to decrease adolescents’ and young adults’ substance use is to alter their perceptions that their friends are using substances.

From one developmental period to the next, we found more evidence for peer selection than influence effects, though. That is, adolescents who reported more substance use at one point in time were likely to report that their peers used more substances at the next point in time, but individuals’ perceptions that their peers were using substances at one point in time were not as consistently related to individuals’ reports of their own substance use at the next point in time. These findings suggest that adolescents who were already using substances were likely to choose to spend time with peers who also used substances.

However, because the time lags in this study covered years, the findings cannot shed light on how peer selection and influence operate from moment to moment. For example, in real time, if adolescents are at a party where a peer offers a beer or a joint, that moment likely increases adolescents’ substance use. Adolescents can also choose activities in which substance use is more likely (such as an unsupervised party) or not as likely (such as a school-sponsored event).

Overall, our findings suggest three main conclusions:

• Peer selection may play a larger role than peer influence in the development of substance use over long developmental periods.
• Even after accounting for stability in individuals’ own substance use from early adolescence to adulthood, perceptions of peers’ substance use in early adulthood predicted adults’ substance use in the fourth decade of life, particularly in the lower-risk community sample.
• Perceptions of peers’ substance use predicted not only alcohol and marijuana use but also opioid and other substance use in the lower-risk sample, suggesting that disrupting alcohol, marijuana, and other substance use early in life has the potential to disrupt potentially more serious substance misuse in adulthood.

LinkedIn image: Motortion Films/Shutterstock. Facebook image: pikselstock/Shutterstock