Abstract
Background: University students often experience significant stress, which may contribute to disordered eating behaviors. Objectives: This study examines the relationship between perceived stress and eating attitudes among physical therapy students in Romania, exploring the impact on overall well-being and identifying predictors of maladaptive eating patterns. Methods: A cross-sectional survey was conducted with 192 students from the Faculty of Midwifery and Nursing, "Carol Davila" University of Medicine and Pharmacy. Two validated instruments were used: the Eating Attitudes Test (EAT-26) to assess eating behaviors and the Perceived Stress Scale (PSS-14) to measure stress levels. Statistical analyses included correlations, ANOVA, and regression models, with a significance threshold of p < 0.05. Results: Among respondents, 15.1% displayed disordered eating behaviors, with stress significantly correlating with EAT-26 subscales. Stress levels were predominantly moderate to high, with younger students reporting higher perceived stress. A significant positive correlation was found between perceived stress and disordered eating behaviors, particularly dieting and bulimia/food preoccupation. Female students reported higher stress levels than males; however, no significant gender differences were observed in disordered eating patterns. No significant associations were found between BMI, weight fluctuations, and EAT-26 scores, suggesting other factors may better explain disordered eating behaviors. Logistic regression identified perceived stress as a significant predictor of disordered eating risk, reinforcing its role in shaping maladaptive eating patterns. Conclusions: Stress significantly impacts eating attitudes, highlighting the need for targeted interventions to reduce stress and promote healthy coping mechanisms. Future research should investigate longitudinal patterns and the effectiveness of interventions aimed at improving student well-being. The limited sample size and the absence of sensitivity analyses are notable limitations that constrain the generalizability of the findings.