Self-Compassion Protects Against Burnout in Medical Students: A Mediation Analysis of Perceived Stress

Objectives This study aimed to examine whether perceived stress mediates the association between self-compassion and burnout in medical students. Methods A cross-sectional study was conducted using an online questionnaire. A total of 649 medical students from a public university in Egypt were recruited via convenience sampling. Participants completed the Self-Compassion Scale–Short Form (SCS-SF), the Perceived Stress Scale (PSS-10), and the Burnout Assessment Tool (BAT-23). Bivariate associations were tested with Pearson correlations; multivariable linear regression assessed predictors of burnout. Mediation analysis was performed with bias-corrected bootstrapped confidence intervals (10,000 samples) to estimate indirect effects. Results Sample means were 36.4 (SCS-SF), 22.2 (PSS-10), and 3.38 (BAT-23). Bivariate analyses revealed significant negative correlations between self-compassion and both perceived stress (r = -0.577) and burnout (r = -0.478), and a positive correlation between stress and burnout (r = 0.603; all p-values < .001). The regression model explained 39.9% of burnout variance (R² = .399); perceived stress was the strongest predictor (β = 0.488), while self-compassion remained a significant protective factor (β = -0.196). The mediation analysis confirmed a significant indirect effect of self-compassion on burnout via perceived stress (B = -0.026, 95% CI [-0.032, -0.021]). As the direct effect also remained significant (B = -0.018), a partial mediation model was supported, with the indirect pathway accounting for 59.2% of the total effect. Conclusions Higher self-compassion is associated with lower burnout largely through reduced perceived stress. Training self-compassion alongside stress-reduction strategies may help safeguard medical student well-being.