How does attending a better school affect cannabis misuse in adolescents?

Dudovitz RN, Chung PJ, Reber S, Kennedy D, Tucker JS, Shoptaw S, Dosanjh KK, Wong MD.  Assessment of exposure to high-performing schools and risk of adolescent substance use: a natural experiment. JAMA Pediatr 172(12):1135-1144

Accompanying editorial: Liljenquist KS, Coker TR. Addressing risky health behaviors among vulnerable youth through the school environment. Location, location, location. JAMA Pediatric 2018;172(12):1119-1120

Why was this study conducted?

Dudovitz et al. wanted to determine whether winning a lottery to attend a high-performing charter secondary school (the intervention group) or attending another school (control group) affected the health behaviours of low-income, minority (mostly Latino) students in Los Angeles.  Self-reported data were collected in the 8^th (just prior to entering secondary school), 9^th, 10^th and 11^th grade on health behaviours (marijuana and alcohol use and misuse, fighting, delinquency and sexual activity) and school-related outcomes.

What does this study add?

From the 9^th grade on, when controlling for covariates such as sex, race/ethnicity, grade point average, family structure, etc., there were trends for students attending high-performance charter schools to report less marijuana use, marijuana misuse and marijuana use by peers than those in the control group. Overall, the only comparison between intervention and control students that was statistically significant was for marijuana misuse score (0.46 for the intervention group vs. 0.71 for the control group, p<.05). When analyzed by sex, the effect of attending a high-performance charter school appeared to be stronger for boys than girls.

Is there anything else I should know?

This is a quasi-experimental study as the researchers had no control over the allocation of students to schools.  It is not clear whether results can be generalized outside of an urban centre with a high proportion of minority students and a number of independently-operated charter schools.  As noted in the accompanying editorial (Liljenquist and Coker), the school environment provides a unique and perhaps powerful opportunity to influence adolescent health behaviours.  At the same time, school environment, youth, family and other social institutions affect health behaviours in complex ways.  A more detailed understanding of all these relationships and inter-relationships would be helpful.