**Background:** University students are a vulnerable population with high prevalence of depressive symptoms (30.6% worldwide) and anxiety (24.5%). Academic burnout, intolerance of uncertainty, and learned helplessness are associated with poor mental health and academic difficulties. The COVID-19 pandemic exacerbated these issues. Existing interventions often neglect learning strategies and intolerance of uncertainty, and face barriers such as stigma and low uptake. This study tested an innovative program addressing these gaps in both online and face-to-face formats.
**Methods:** 105 French university students were divided into three groups: online intervention (n=36, 94.4% female, mean age 19.3±1.6), face-to-face intervention (n=32, 93.7% female, mean age 19.5±2.2), and control (n=37, 78.3% female, mean age 19.4±2.3). The 9-week program covered stress, emotions, emotion regulation, learning, learning strategies, motivation, nutrition, sleep, and worry via identical 10-minute videos. The online group watched videos on YouTube and interacted via Discord; the face-to-face group watched in class with group exercises and discussions. Assessments at baseline (T0) and 10 weeks later (T1) used the Hospital Anxiety and Depression Scale, Maslach Burnout Inventory-General Student Survey, Intolerance of Uncertainty Scale-Short Form, Learned Helplessness Questionnaire, Social Provisions Scale, and visual analog scale items for learning strategies. Nonparametric within-group comparisons (Wilcoxon tests) were conducted.
**Key Results:** In the online intervention group, anxiety decreased slightly (T0 M=9.6, SD=4.1 vs T1 M=8.6, SD=3.8; W=387.5, p=.05, r_rb=0.16). No significant change in depressive symptoms or total academic burnout was observed in any group. Academic inefficacy decreased in the face-to-face group (T0 M=19.3, SD=8.7 vs T1 M=13.9, SD=6.2; W=334.0, p=.03, r_rb=0.26). Intolerance of uncertainty decreased in both the online group (T0 M=37.2, SD=10.5 vs T1 M=34.4, SD=9.2; W=397.0, p=.01, r_rb=0.33) and face-to-face group (T0 M=40.4, SD=11.1 vs T1 M=35.2, SD=11.8; W=325.0, p=.02, r_rb=0.31), with no change in controls. Learned helplessness decreased in both the online group (T0 M=26.9, SD=9.2 vs T1 M=24.5, SD=7.4; W=326, p=.05, r_rb=-0.02) and face-to-face group (T0 M=27.5, SD=7.3 vs T1 M=24.3, SD=6.2; W=438.0, p=.001, r_rb=0.65), with no change in controls. Positive attitude toward errors increased in both intervention groups (online: W=97.0, p=.04, r_rb=-0.48; face-to-face: W=133.0, p=.04, r_rb=-0.46). Planning of learning sessions increased only in the face-to-face group (T0 M=66.1, SD=34.9 vs T1 M=75.3, SD=30.4; W=10.0, p=.04, r_rb=-0.71). Perceived ease in seeking help increased only in the face-to-face group (T0 M=42.1, SD=30.3 vs T1 M=66.7, SD=26.6; W=39.0, p<.001, r_rb=-0.84). Perceived social support increased only in the face-to-face group (T0 M=34.1, SD=4.9 vs T1 M=35.4, SD=4.3; W=62.0, p=.03, r_rb=-0.75). The online group reported more deep learning strategies (self-management: T0 M=0.39, SD=0.83 vs T1 M=0.61, SD=0.81; W=19.5, p=.05, r_rb=-0.89) and the face-to-face group used more deep strategies (self-management: T0 M=0.18, SD=0.39 vs T1 M=0.59, SD=0.61; W=14, p=.005, r_rb=-0.94) and fewer surface strategies (rehearsal: T0 M=2.0, SD=1.24 vs T1 M=1.59, SD=1.01; W=155.5, p=.05, r_rb=-0.41).
**Clinical Implications:** Both program formats reduced intolerance of uncertainty and learned helplessness, key risk factors for anxiety and burnout. The face-to-face format additionally improved perceived social support, academic self-efficacy, help-seeking, and reduced surface learning strategies, suggesting that embedding interventions in classroom teaching may yield broader benefits. The online format showed a slight reduction in anxiety. These findings support implementing such programs in universities, with a preference for face-to-face delivery when possible. Limitations include small sample size, high dropout rate (especially online), lack of randomization, and short follow-up.