Randomized Trial of a Social Networks Intervention

Substance use is an increasing problem among adolescents. Despite decades of prevention work, most recent data suggest that, by the end of high school, a quarter of teens regularly use marijuana and nearly half regularly use alcohol. While adolescent marijuana use has increased across the country in the last decade, rates in California are higher than the national average. This may be partially attributable to increased social acceptance and access through legalized dispensaries. As the movement to legalize marijuana gains traction, similar changes are likely to occur throughout the country. The high prevalence of alcohol and marijuana use is particularly disturbing in light of evidence that even intermittent adolescent exposure may have long-term impacts on brain development, in addition to associations with other risky health behaviors and future substance abuse disorders. African American and Latino youth are especially at risk, making adolescent substance use prevention a potentially important strategy to address health disparities.

Few effective and sustainable substance use interventions exist for high school students. Preventing substance use becomes more challenging as adolescents age, with fewer prevention programs demonstrating success for high school youth than for middle school youth. The challenge is to identify modifiable predictors of substance use for high-school youth and harness them for prevention. Furthermore, most interventions are difficult to disseminate and sustain due to limits on financial and human capital. AVID is already widely implemented in public schools. Thus, if AVID is effective in reducing substance use while also improving academic outcomes, the program would serve as a promising model for future interventions.

Social networks are likely to powerfully influence adolescent substance use. There is strong and consistent evidence that adolescent substance use is closely tied to the behaviors and attitudes of individuals in their social network. Social network structure and composition are shaped by cultural and environmental factors. In school environments, these factors include policies like tracking of students based on prior achievement. Social networks then determine sources of support, transmission of social norms, and access and opportunity to engage in substance use. Although peers are powerful points of influence during adolescence, relationships with supportive adults can continue to protect against substance use.

Despite great interest in social networks as potential determinants of risky behaviors, causal pathways linking social networks with adolescent health have yet to be fully categorized. Studies suggest that changes in social networks can impact substance use, but little is known about how to harness this relationship effectively. Answering this question may have applications across many social environments and health domains. Interventions that re-wire social networks might offer a relatively low-cost, self-sustaining, and effective means to prevent some of the most challenging behaviors that lead to poor health outcomes. School environments present a critical opportunity to impact adolescent health. Through near-daily exposure, school environments have the potential to shape adolescents' relationships, social norms, sense of identity, and educational attainment. These factors are likely to be mutually reinforcing and are strongly associated with substance use.

AVID, a widely used educational intervention, offers an important opportunity to test whether schools can re-wire social networks to improve adolescent health. AVID removes middle-performing students from a typical school environment, where social pressure often reinforces low academic achievement and problem behaviors, and places them alongside high-performing students in a college-preparatory environment where positive social norms are valued. AVID capitalizes on a natural transition point, when social networks are in flux, by targeting students at the beginning of 9th grade. Perhaps not coincidentally, 9th grade is also an enormously vulnerable period-36% of high school drop-outs do so in 9th grade, and only 15% of 9th graders who are held back will ever graduate. Finally, the program emphasizes the student/teacher relationship, training teachers to serve as mentors and student advocates, providing both academic and emotional support to cultivate a family-like atmosphere. Through the proposed randomized controlled pilot evaluation of AVID, investigators can measure whether the program successfully changes social networks, and test whether changes in social networks motivate changes in substance use. Understanding whether AVID's return on investment extends to health outcomes is vitally important given the limited resources available to support multiple competing K-12 education initiatives.

The Los Angeles Unified School District (LAUSD) implemented AVID to provide academic and social support to students from groups traditionally under-represented in higher education (e.g., students of color, first generation to attend college). AVID is an international non-profit institution originally established in 1980 by a San Diego high school teacher. AVID's goal is to prepare vulnerable students functioning in the academic middle for success at a 4-year university. Students in each school must actively apply to the program to demonstrate both eligibility and interest. AVID's eligibility criteria include having a grade point average of 2.0-3.5 and returning a contract signed by the student and a parent indicating that the student chooses to participate in the program, has a desire to go to college, and agrees to enroll in rigorous academic courses. Identification of eligible candidates for AVID occurs during the second half of 8th grade, prior to high school entry, and is led by the AVID coordinator. At nearly every AVID high school, the number of students who qualify for AVID far outweighs the program's capacity-there are, on average, 40 AVID slots per school, which results in AVID serving only 12% of the student body. In LAUSD, the demographics of AVID students match those of the district-low-income, and African American or Latino. For example, in 2014, 86% of LAUSD's AVID students were Latino, 6% were African American, and 89% qualified for free or reduced lunch.

LAUSD's AVID program has agreed to allow investigators to recruit study participants from the pool of students who meet the AVID eligibility requirements as well as a group of high performing (middle school grade point average (GPA) >3.5) students who are not AVID eligible. Near the end of 8th grade, LAUSD will randomize all AVID eligible applicants through a lottery into either AVID or a waitlist group. Prior to the random assignments being revealed, investigators will ask all students randomized into AVID and an equal number of students randomized into the waitlist group to return a signed parental consent form for study participation and to complete a survey. Students will be recruited and surveyed prior to revealing their assignment to avoid any response biases caused by happiness or disappointment at the lottery result. Decisions to participate in the study will have no impact on the lottery itself. Investigators will survey all study participants at the end of 8th grade, 9th grade, 10th grade, and 11th grade.