A Nature-Based Intervention and Mental Health of Schoolchildren: A Cluster Randomized Clinical Trial

Abstract

Importance: Nature-based therapeutic or preventive interventions for mental health are increasingly popular, but their effectiveness for improving mental health is not well documented.

Objective: To investigate the effectiveness of the Open Sky School Program (École à Ciel Ouvert), a 12-week nature-based intervention for elementary schoolchildren in grades 5 and 6, for reducing mental health symptoms.

Design, setting, and participants: This 2-arm, cluster randomized clinical trial was conducted from February 27 to June 16, 2023, in French-language elementary schools in Quebec, Canada, with green space within 1 km. Participants were teachers and students in grades 5 and 6.

Intervention: In the intervention group, for 2 hours per week for 12 weeks, classes were taught outdoors in a nearby park or wooded area. Teachers were encouraged to engage students in basic subjects and in 10 or more mental health activities (mindfulness, philosophy, and/or art therapy). Control group schools conducted classes as usual.

Main outcomes and measures: The primary outcome was change in student mental health (internalizing and externalizing symptoms, social problems) based on teacher- and student-reported 30-item Social Behavior Questionnaire (SBQ) scores (3-point scale) from baseline to the immediate postintervention follow-up, assessed in per-protocol and intent-to-treat mixed-model analyses. Secondary outcomes were student self-reported changes in depressive symptoms, positive or negative affect, pro-environmental efforts and/or attitudes, and nature connectedness.

Results: A total of 33 schools participated (53 teachers, 1015 students), including 16 schools (25 teachers, 515 students) in the intervention group and 17 schools (28 teachers, 500 students) in the control group. Student mean (SD) age was 10.9 (0.75) years; 507 (50.7%) were girls. Per-protocol and intent-to-treat mixed-model analyses showed no differences in mental health symptom change between groups; for example, the adjusted mean difference in SBQ scores between the intervention and control groups for externalizing symptoms was -0.04 (95% CI, -0.13 to 0.04) in the intent-to-treat analysis and -0.06 (95% CI, -0.16 to 0.04) in the per-protocol analysis. Post hoc analyses revealed low mental health symptoms at baseline, with low variability. Slightly greater reductions in symptoms were observed in the intervention group, but only for children with higher mental health symptoms at baseline (P < .05 for interaction). For example, for children with internalizing symptoms 1 SD above the mean at baseline, internalizing symptoms decreased by 0.38 SD (mean change, -0.15; P < .001) in the intervention group vs the control group.

Conclusions and relevance: In this large cluster randomized clinical trial in daily-life elementary school settings, 12 weeks of classes in green space for 2 hours per week did not reduce mental health symptoms in students aged 10 to 12 years in either the per-protocol or the intent-to-treat analysis. However, this low-cost, safe outdoor intervention may provide unmeasured or longer-term benefits for children with higher risk of mental health symptoms.

Trial registration: ClinicalTrials.gov Identifier: NCT05662436.

Figure 1.. CONSORT Flow Diagram of Study Participants

Figure 2.. Mental Health Symptoms as Reported by Teachers at Postintervention Assessment as a Function of Mental Health Symptoms at Baseline in the Intervention and Control Groups

Symptoms were assessed using the Social Behavior Questionnaire (SBQ). The SBQ evaluates frequency of symptoms over the past 2 months on a 3-point scale (never or not true = 0, sometimes or somewhat true = 1, often or very true = 2). We retained a continuous score for each subscale. Results were adjusted for student sex, age, birth outside Canada, and primary language other than English or French; teacher born outside Canada and teacher’s years of teaching experience; and school socioeconomic disadvantage indicator and Normalized Difference Vegetation Index (quantified by school postal code within a 250-m buffer zone).