- An initiative in Colombia examines how early detection of depression and anxiety can help prevent school violence and strengthen adolescents' educational trajectories.
- The study found that nearly 19,000 adolescent students in Medellín have high levels of depression and anxiety, with impacts linked to risks of school dropout, violence, and social exclusion.
- The findings show that mental health is closely tied to decisions that shape young people's futures, including whether they stay in school and their exposure to risky behaviors.
Can a teenager's mental health determine whether they finish school, get drawn into a a criminal gang or enter the sex trade?
Around ages 13 and 14, mental health can play a pivotal role in how adolescents respond to opportunities, risks, and pressures that shape their future. School is the only place where the State reaches nearly every adolescent during this window. That makes the classroom a natural point for preventing violence before it happens. Unlike other institutions, school can offer a student timely support without depending on the family to seek help: the student is already there every day, which makes it the point of greatest reach and lowest cost for timely intervention.
This post is part of Schools that Care, a series of posts from the Inter-American Development Bank (IDB) that brings together evidence and experience on how education systems can help prevent violence and build safer environments for children and adolescents. From Medellín, Colombia, this piece explores a fundamental dimension of that agenda: mental health and schools' potential to identify risks early and offer support before violence, dropout, or exclusion take hold.
Together with Medellin's Secretariat of Education, we tested that idea with a research team from Universidad EAFIT, the University of Chicago, and Innovations for Poverty Action (IPA). Between 2025 and 2026, we surveyed nearly 19,000 middle school students in grades 7 and 8 at 135 public schools in the city, measured symptoms of depression and anxiety using internationally validated instruments (the Patient Health Questionnaire for Adolescents PHQ-9), and explored how students imagine their future working lives.
The findings describe a large-scale problem. 42% of students report scores indicating moderate or higher depression, and 29% report anxiety in that same range. Girls' average depression score exceeds boys' by more than three points (on a scale of 0 to 27). This reality extends across the entire education system: we found that less than 2% of the variation is explained by which school a student attends. The remaining 98% reflects differences among students who share the same school, and this holds true across every school in the system. That is why the response must be universal and reach every school. For decision-makers, this means that interventions targeting only a handful of high-risk schools are not enough: public policy must be designed to reach the entire education system.
Description: On average, girls report higher levels of depression than boys, and they account for the most severe cases.
The survey also asked about suicidal ideation. This indicator matters because it allows timely identification of students at greatest risk and can trigger support pathways before a crisis leads to dropping out or violence. Forty percent of students reported some level of suicidal ideation in the past two weeks, and nearly 12% reported thinking about death as a solution almost every day. The burden falls more heavily on girls: about half of them report some level of suicidal ideation, compared with nearly 30% of boys. These reports are not a diagnosis, but they point to the need for institutional follow-up and support pathways within schools.
Suicidal ideation is also more common among girls: about half of girls report some level, compared with nearly 30% of boys.
Students with higher depression report more antisocial behavior among their peers, greater risk of dropping out, greater exposure to substances, and greater interest in risky occupations. Among boys, this includes joining neighborhood gangs or organized crime groups, known locally as “combos”; among girls this includes entering the sex trade. To measure these expectations, we used an indirect method:
- We asked each student to imagine ten young people like themselves—of the same age, neighborhood, and family—and to distribute them across four occupational groups: low-skilled jobs, technical trades, professional careers, and risky activities.
- We then repeated the exercise, removing the legitimate options one by one.
- As those options disappeared, students with more depressive symptoms assigned more young people to the risky activity.
These results show correlations, but they do not prove that depression pushes a young person toward a combo or sex work. The survey captures a single point in time and does not establish causation. Even so, the pattern is consistent and shows that mental health is tied to the conditions under which an adolescent decides about their future.
As legitimate occupational options are reduced, students with more depressive symptoms—especially boys—show a greater tilt toward risky activities.
The Secretariat of Education of the Mayor's Office of Medellín, with support from the Fundación Vibra en Alta and the Universidad de Antioquia, is implementing an intervention based on cognitive behavioral therapy (CBT), the approach with the strongest empirical support for treating depression and anxiety in adolescents. Cognitive behavioral therapy teaches students to recognize thoughts that harm them, question those thoughts, and build tools to manage emotions and everyday problems.
The intervention began as a pilot in 2025. Program psychologists worked with 333 students in 28 schools. Students' self-reported mood at the start of each session improved steadily over the course of the program. In 2026, the program is being expanded to reach nearly 500 students, with weekly psychosocial support and supervision and a referral pathway to specialized care for the most severe cases.
This initiative is designed to generate evidence. We will measure its effect through a rigorous evaluation that assigns support randomly and allows us to estimate its real impact on mental health and, over time, on school retention and student trajectories. In Latin America, there are only a very limited number of studies linking adolescent mental health treatment to long-term educational and economic outcomes. Medellín can help fill that gap. As a complement to in-person support, we are also exploring digital support tools.
The instruments we use are brief, validated, and can be administered by non-clinical staff, which makes it possible to assess an entire student population at low cost and repeat the measurement every year. On that basis, schools can identify students who need support in time and connect mental health to the retention and prevention programs already operating in the city. The measurement also makes it possible to offer differentiated support to girls and boys, who enter adolescence carrying different risk burdens.
Early violence prevention begins with institutions that stand by young people as they make their first major decisions, and schools are often the first of those institutions. Medellín is testing that model and measuring whether it works. Other cities in Latin America can replicate the school-based measurement and random assignment that Medellín is using to test the model. Ultimately, it is rigorous evidence—not intuition—that should determine which programs get scaled up and which are set aside across the region.