Background: Peer influence has been reported as a predominant risk factor of harmful substance use among young people. Given the high prevalence of substance use among young people in Ghana, an understanding of young people’s social networks will be central to developing and implementing preventive interventions. This study described the social network features and substance use prevalence of senior high school students in Ghana, and the underlying mechanisms that may influence substance use behaviour within friendship networks.

Methods: Using a mixed methods study design, we conducted a social network analysis among senior high school students in two schools in a Ghanaian municipality. We used a bounded social network approach and implemented a social network survey as the first phase of data collection process. A second phase consisted of in-depth interviews with a sub-sample of the survey participants. Quantitative data were analyzed using descriptive statistics and inferential statistics. We analyzed qualitative data using a thematic framework approach.

Findings: Quantitative findings revealed a relatively high prevalence of any substance use among the participants (63.5%, 44.7%, and 64.1% in networks 1, 2, and 3 respectively). Also, we found a relatively low friendship network density and average level of reciprocated ties. Gender identity was identified as a strong predictor of friendship network formation. The qualitative findings revealed that the substance use values and norms of a friendship network, as well as traditional gender norms influence individuals’ substance use behaviour.

Conclusion: The study provides findings on the association between young people’s social network features, gender norms, and their substance use behaviour. The findings can be leveraged for the design and implementation of school-based interventions aimed at preventing substance use among young people in Ghana.

Previous literature consistently demonstrates that American Indian (AI) youth have earlier initiation and heavier use of substances compared to other racial groups. This study analyzes the social networks of Alaska Native (AN) youth, focusing on the role of school, family, culture, and peer influences on substance use, depression, and suicide. This cross-sectional analysis included 150 seventh to twelfth grade students enrolled in a charter school in Fairbanks, Alaska. In addition to analyzing the role of peer networks, this study incorporates students’ relationship to school staff. Students were first provided with a school roster and asked to identify their friends. They were next provided with a roster of faculty and selected whom they would go to for support and ranked how likely they were to seek their support. Reciprocal ties between peers were protective against substance use. Intriguingly, having more staff in a student’s network and stronger relationships with those staff were protective against substance use. These novel findings, particularly concerning the influence of staff, provide a new dimension to our current understanding of substance use in AN youth.

This analysis illustrates the complexities of social influence governing substance use while considering other common predictors of substance use among youth, such as depression and suicide. It highlights the need to consider both risk and protective factors within social networks, laying the groundwork for well-informed interventions. The results support harnessing the positive influence of peer networks and staff relationships to reduce substance use in adolescents, presenting an optimistic perspective for community-based interventions.

The rising use of marijuana among young adults has drawn increasing attention from researchers, leading to a surge in studies examining its effects, risks, and patterns of consumption. However, much of this research focuses on marijuana in isolation, overlooking its role within the broader network of substance use. A key question remains unanswered: Is marijuana replacing other drugs, strengthening connections between substances, or creating a more distinct pattern of use? Understanding these shifts is crucial, as substance use rarely occurs in isolation but rather within an interconnected system. Traditional analyses fail to capture these dynamics, highlighting the need for a network-based approach to better inform intervention strategies.

In this study, we used repeated national cross-sectional surveys of university students to examine the evolving structure of drug use networks. Using mixed graphical models, we constructed substance use networks at each time point, analyzing global metrics, such as density, average path length, clustering coefficient, and modularity, to assess shifts in connectivity and community structure. Additionally, we explored node-level measures, including weighted degree, betweenness, closeness, and eigenvector centrality, to determine how patterns of co-use have changed over time. Our primary aim was to assess whether marijuana use has not only increased but also whether its position in the substance use network has shifted, indicating changes in its role within polysubstance patterns.

We used repeated cross-sectional survey data from Colombian university students (2009: n = 7,803; 2012: n = 9,876; 2016: n = 9,805; 2023: n = 11,065), collected by the Colombian government. Mixed Graphical Models were used to estimate conditional dependencies between substances at each time point. Global network metrics (density, average path length, clustering coefficient, modularity) were computed to assess overall connectivity and community structure, while node-level measures (weighted degree, betweenness, closeness, eigenvector centrality) were used to examine marijuana’s changing role within the network. Networks were visualized using ggraph in R.

Descriptive results indicate a decline in lifetime prevalence of alcohol and tobacco use, while the use of ecstasy, tranquilizers, and marijuana has increased. Network analyses reveal that overall connectivity has remained high (density = 0.86–0.89), while the network has become more compact over time, as indicated by a decreasing average path length (from 0.53 in 2009 to 0.36 in 2023). At the substance level, partial correlations indicate that marijuana’s connections with other drugs have strengthened in recent years, reflecting its increasing prevalence. Centrality measures further highlight marijuana’s growing role within the network, even as alcohol and tobacco continue to serve as foundational substances in polysubstance use.

These findings suggest that while the overall substance use network remains highly interconnected, marijuana has become a more integral component of polysubstance patterns. This shift underscores the need for prevention and intervention strategies that account for marijuana’s evolving role in the broader drug use landscape.

Adolescent social networks have been shown to be strongly associate with use of tobacco, marijuana, and alcohol. In this study we replicate those associations and make four additional contributions to the literature. First, we show that both perceived friend use and friend self-report are independently related to associated with individual use. Second, adolescents who named more friends in response to a network question were less likely to report substance use, whereas those who were named as friends were no more or less likely to report use. Third, students who identified as Asian ethnicity were less likely to report any substance use but remained susceptible to peer influence on substance use. Finally, we show that being named as a friend by Asian students was protective against use of all three substances. Data were collected from up to 14 demographically diverse schools in southern California and consist of 23,012 surveys collected bi-annually from 9th to 12th grade.